You are on page 1of 67

Performance

Evaluation Report

Indonesia: Decentralized
Health Services Project

Independent

Evaluation

Performance Evaluation Report


December 2013

Decentralized Health Services Project in


Indonesia

Reference Number: PPE: INO 2013-21


Loan Number: 1810-INO
Independent Evaluation: PE-769

Note
(i)
(ii)

The fiscal year (FY) of the Government of Indonesia ends on


31 December.
In this report, $ refers to US dollars.
Key Words

advocacy, asian development bank, capacity building, decentralized health services


project, demographic and health survey family planning services, health insurance,
health sector reforms, health service utilization financing program, health system
management, independent evaluation department, indonesia, infant and maternal
mortality rate performance evaluation report, project management support, under-5
mortality rate

Director
Director General
Director
Team leader
Team members

V. Thomas, Independent Evaluation Department (IED)


W. Kolkma, Independent Evaluation Division 1, IED
G. Rauniyar, Principal Evaluation Specialist, IED
P. Lim, Evaluation Officer, IED
V. Melo-Cabuang, Evaluation Assistant, IED

The guidelines formally adopted by the Independent Evaluation Department (IED) on


avoiding conflict of interest in its independent evaluations were observed in the
preparation of this report. To the knowledge of IED, there were no conflicts of interest
of the persons preparing, reviewing, or approving this report.
In preparing any evaluation report, or by making any designation of or reference to a
particular territory or geographic area in this document, IED does not intend to make
any judgments as to the legal or other status of any territory or area.

Abbreviations
ADB
BAPPENAS

BKKBN

DHS
DHS2
IDHS
IED
IMR
M&E
MDG
MMR
MOH
MNCH
NGO
PCR
SDR
SUSENAS
TA
U5MR

Asian Development Bank


Badan Perencanaan dan Pembangunan Nasional (National
Development Planning Agency)
National Population and Family Planning Board (Badan Koordinasi
Keluarga Berencana Nasional)
Decentralized Health Services Project
Second Decentralized Health Services Project
Indonesia Demographic and Health Survey
Independent Evaluation Department
infant mortality rate
monitoring and evaluation
Millennium Development Goal
maternal mortality ratio
Ministry of Health
maternal, newborn, and child health
nongovernment organization
project completion report
special drawing right
National Socioeconomic Survey (Survei Sosial Ekonomi Nasional)
technical assistance
under-5 mortality rate

Currency Equivalents
Currency unit rupiah (Rp)

Rp1.00
$1.00

=
=

At Appraisal
14 December 2000
$0.00011
Rp9,315

At Project Completion
30 June 2009
$0.000097
Rp10,257

Glossary
Askeskin
Jamkesmas
Jampersal
JPS-BK
polindes
posyandus
puskesmas

health insurance for the poor


health service utilization financing program
maternal health financing program
Social Safety Net Program for Health
village delivery center
village health posts
health centers

At Evaluation
30 November 2013
$0.000085
Rp11,755

Contents
Acknowledgments
i
Acknowledgments
Basic Data ................................................................
................................................................................................
................................................................................................
...................................................................
................................... iii
Executive Summary ................................................................
................................................................................................
......................................................................................
...................................................... v
Chapter 1: Introduction
1
Introduction
A.
Evaluation Purpose and Process ...............................................................................1
B.

Expected Results and Project Objectives ...................................................................2

C.

Project Completion Report Assessment ....................................................................3

Chapter 2: Design and Implementation ................................................................


........................................................................................
........................................................ 4
A.
Formulation..............................................................................................................4
B.

Rationale ..................................................................................................................4

C.

Cost, Financing, and Executing Arrangements .........................................................5

D.

Procurement, Construction, and Scheduling ............................................................7

E.

Design Changes........................................................................................................8

F.

Outputs ....................................................................................................................8

G.

Consultants ............................................................................................................13

H.

Loan Covenants ......................................................................................................14

Chapter 3: Performance Assessment................................................................


............................................................................................
............................................................15
............................15
A.
Overall Assessment .................................................................................................15
B.

Relevance ...............................................................................................................15

C.

Effectiveness ..........................................................................................................19

D.

Efficiency ................................................................................................................23

E.

Sustainability ..........................................................................................................25

Chapter 4: Other Assessments................................................................


................................................................................................
.....................................................................
.....................................28
.....28
A.
Impacts ..................................................................................................................28
B.

Performance of ADB, Borrower and Development Partners ....................................32

C.

Technical Assistance ...............................................................................................35

Chapter 5: Issues, Lessons, and FollowFollow-Up Actions ................................................................


.......................................................................
.......................................37
.......37
A.
Issues .....................................................................................................................37
B.

Lessons ...................................................................................................................38

C.

Follow-up Actions ..................................................................................................41

Appendixes
1. Design and Monitoring Framework ................................................................................44
2. Rating Matrix for Core Evaluation Criteria .......................................................................48
3. Economic Reevaluation ..................................................................................................49
4. Health Outcomes in Project Provinces .............................................................................55

Acknowledgments
This report is a product of the Independent Evaluation Department (IED) of the Asian
Development Bank (ADB). Ganesh Rauniyar, Principal Evaluation Specialist, led the
evaluation with the support from Ma. Patricia Lim and Valerie Anne Melo-Cabuang.
Ross McLeod (health economist and consultant) provided the analytical support for the
study.
The evaluation team appreciates inputs and support provided by the Indonesia Resident
Mission and Ministry of Health staff at headquarters and in the field. The evaluation
team benefited from discussions with staff of the Decentralized Health Services Project
(phase II), the National Development Planning Agency (BAPPENAS), Statistics Indonesia,
and National Family Planning Coordination Board, and with key development
partnersAustralian Agency for International Development, the German Society for
International Cooperation, United Nations Children Fund (UNICEF), United States
Agency for International Development, and the World Health Organizationthat have
lent significant support to Indonesias health sector. Insights from former Ministry of
Health staff and consultants also strengthened the study.
The report was peer-reviewed by Maya Vijayaraghavan and Linda Arthur within IED.
It also benefited from valuable discussions and comments on the draft report from the
Southeast Asia Department of ADB.
The report was prepared under the overall guidance of Vinod Thomas, Director General,
and Walter Kolkma, Director (Division 1), IED. IED retains full responsibility for the
report.

Basic Data
Loan 18101810-INO Decentralized Health Services
Project Preparatory and Institution Building Technical Assistance
TA
No.
3448
3579

TA Name

Type

Decentralized Health Services


Support for Health Sector Policy
Reform

PP
AD

Borrower
Executing Agency

Approval Date

180.00
1,000.00

Per ADB Loan


Documents
87.00
65.00

Key Project Data ($ million)


Total project cost
ADB loan amount/utilization
ADB loan amount/cancellation
Key Dates
Appraisal mission
Inception mission
Loan negotiations
Board approval
Loan agreement
Loan effectiveness
Loan closing
Months (effectiveness to completion)

Amount
($000)

Expected

30 Sep 2006

26 May 2000
14 Dec 2000

Actual
91.57
59.31
15.88
Actual
25 Sep23 Oct 2000
312 Apr 2001
1617 Nov 2000
14 Dec 2000
27 Mar 2001
25 Jun 2001
30 Jun 2009
96

Republic of Indonesia
Ministry of Health

Type of Mission
Appraisal
Inception
Review
Midterm review
Project completion review
Independent evaluation mission

No. of Missions
1
1
13

No. of PersonPerson-Days
49
0
155
54
20
14

AD = advisory, ADB = Asian Development Bank, PP = project preparatory, TA = technical assistance.

Executive Summary
The project performance evaluation report of the Decentralized Health Services
Project in Indonesia provides a set of lessons from a sector-wide, reform-based health
program in a country going through a process of decentralizing government services
since 1999. The findings from the evaluation will feed into higher-level evaluations,
including validation of the country partnership strategy final review planned for 2014.
The Asian Development Bank (ADB) approved a loan of $65 million and project
preparatory technical assistance (TA) of $800,000 on 14 December 2000. The loan
closed on 30 June 2009 after two extensions. An advisory TA of $1 million, financing
36 person-months of the work of an individual consultant, was attached to the loan
with the objective of strengthening local health sector reform. The project was
originally designed to cover 50 districts and 12 cities in seven provinces but,
as decentralization evolved, the coverage expanded to 73 districts in eight provinces.
The project aimed to improve the health status of the population in the project
areas. It had twin objectives of (i) improving health and family planning services, and
(ii) guaranteeing access of the poor to essential health and family planning services.
Project components comprised (i) advocacy and capacity building; (ii) adaptation of
health services to local needs through appropriate health sector reforms;
(iii) investments in health and family planning; and (iv) project management support
(e.g., implementation, and monitoring and evaluation). Likewise, the advisory TA
envisaged six outputs: (i) health sector reforms identified, (ii) capacity for health system
management and health service delivery built, (iii) health plans developed,
(iv) awareness of health priorities created among decision makers, (v) operations
research capacity developed, and (vi) project management supported.
The evaluation concludes that project performance was successful overall,
which was also the project completion reports assessment. Primary, district, and
provincial health services have developed to improve access for a greater number of
people, including those in remote and rural areas, because of project support.
Locally, however, health service delivery has been constrained by limited
funding that focuses on administrative expenditure and involves significant transaction
costs. Insufficient recurrent expenditure, limited hospital beds and specialists, and high
staff turnover are often inadequately dealt with. Nevertheless, the central government
is responding to some of these concerns with the introduction and subsequent scaling
up of health insurance schemes such as those initiated under the project.
The project is rated relevant at the time of design, during implementation,
at completion, and at evaluation. It was also relevant in the context of Indonesias
development priorities, ADB's country strategy, and the Millennium Development Goals
associated with child and maternal health. The project design would have benefited
from smaller coverage (fewer provinces) and a more rigorous process in targeting
disadvantaged populations to further enhance project relevance. Greater consideration
of project risks, particularly related to institutional capacity and political economy, and
development of risk mitigation strategies would have added rigor to the project design.
The evaluation team notes that the midterm review of the project led to greater focus

vi

Decentralized Health Services Project in Indonesia


on maternal and child health. This resulted in less emphasis on family planning and
nutrition.
The project's performance is rated effective in achieving the stated outputs and
outcomes. The health service infrastructure built or renovated with project support was
found to be in good condition and well maintained. Similarly, health facilities are using
the project-supported equipment well. A large number of health service staff received
management and technical training and fellowships for higher studies. Some of them
now occupy key positions in the health system. A review of health facility utilization
suggests that poor households who otherwise could not afford health services have
benefited the most. Health insurance coverage for the poor has also been a major
catalyst for greater use of health facilities. Some of the reform measures introduced by
the project, particularly subsidy to trained birth attendants and health insurance for the
poor, have been scaled up and are likely to benefit disadvantaged groups.
Overall, the health outcome indicators show improvements across Indonesia,
although no significant differences are observed between project-supported and nonproject-supported districts or provinces. This applies to both the infant mortality rate
and maternal mortality ratio. As suggested earlier, the project set ambitious targets and
geographical coverage, which resulted in resources being spread thinly. Some of the
targets proved unrealistic. For example, the project aimed to have at least 90% of
deliveries assisted by trained birth attendants. This has been achieved only in Bali.
In Central Sulawesi and Southeast Sulawesi coverage of skilled delivery is still below
70%. The size of resources committed across the vast implementation area was too
small to make an impact on the provinces as a whole, even though the project led to
positive results locally.
The project is rated less efficient due to implementation delays and limited
economic returns. A six-month delay in loan effectiveness, two extensions to the
project implementation period, and late release of counterpart funds contributed to
this. Although the project produced most of its planned outputs, outcomes have not
fully materialized. Most notably the revision of the economic analysis conducted during
appraisal indicates that the project impacts have been significantly smaller than
anticipated. The analysis at appraisal was based on the burden of disease in Indonesia
being reduced by 4% compared with the burden of disease in developed countries.
Given that the difference in health indicators in project and non-project provinces has
not changed significantly, a 4% reduction in burden of disease is unlikely to have
happened.
The evaluation finds that the project benefits are likely sustainable. Civil works
and equipment procured under the project are fully utilized and many of the staff
trained under the project have moved to key positions in local health administrations.
Project-supported policy reform initiatives such as health insurance piloting in districts
of Bali have been rolled out across the province, and are likely to be replicated as part
of the upcoming expansion in the national insurance scheme. Schemes to subsidize
traditional birth attendants are still being implemented and training modules are now
included in local government curriculums and training programs. Furthermore, local
governments have demonstrated commitment to improving health service delivery and
increasingly began allocating more funds for health interventions. In cities, a number of
well-equipped hospitals have been established and are available for referral services
when required.

Executive Summary

The evaluation concurs with the project completion report's assessment that
ADB's performance was satisfactory. ADB fielded missions on time and accommodated
government requirements, including two extensions of the loan closing date.
Government performance, on the other hand, is rated less than satisfactory. Delayed
release of counterpart funds of up to 6 months in a given year limited project activities
that could only be taken up once funds were released. Project funds could have been
used in fewer provinces with higher poverty incidence and lagging health indicators.
Local area investment planning also should have been more pro-poor in its resource
allocation. A more explicit guideline should have been issued on planning project
allocations for different tiers of the health system and targeting lower-income districts.
Efforts to encourage local governments to increase recurrent health expenditure should
have been more emphatic.
Indonesia continues to face hurdles in the health sector. Key challenges are:
(i)
Needs are greatest in remote and rural areas. In many areas, outreach is
limited due to geographical isolation.
(ii)
Wide disparities in technical and managerial capabilities, as well as in
monitoring systems, remain across health service centers at all levels.
(iii)
The local health system puts too much emphasis on administrative
support.
The evaluation recommends four measures:
(i)
ADB should remain engaged with the Government of Indonesia in the
health sector because a large segment of the population is still in dire
need of modest health services. This is consistent with ADBs inclusivegrowth approach.
(ii)
ADB should promote that local, district, provincial, and central
governments work on and adhere to a clear funding framework for
stronger health and family planning services at all levels. Health service
facilities need reliability and less complexity in funding for
infrastructure development, procurement of equipment and drugs, and
retention of qualified health workers.
(iii)
ADB should help the provincial governments, with support from the
central government, to promote twinning arrangements with health
facilities in other provinces and cities. This can serve as a sustainable
model for increasing outreach of health services as well as skills
development of health workers.
(iv)
ADB should have policy dialogue with the central government
regarding the need for a lead role of the National Development
Planning Agency, Ministry of Health, National Family Planning
Coordination Board, and Statistics Indonesia in ensuring that healthrelated statistics are collected regularly and records are updated in a
timely manner.

vii

CHAPTER 1

Introduction
1.
This chapter describes purpose and process adopted for the project
performance evaluation. It states expected results and outlines project objectives and
highlights key assessments provided in the project completion report.

A.

Evaluation Purpose and Process

2.
The Decentralized Health Services Project (DHS) in Indonesia was selected for
evaluation to determine lessons from a sector-wide, reform-based program in a country
experiencing decentralization. It also aimed to fill the knowledge gap in the
performance of health projects supported by the Asian Development Bank (ADB).
The evaluation findings will feed into other thematic and country evaluations by the
Independent Evaluation Department (IED) in the near future. IED did not validate the
findings of the project completion report (PCR) for this project.1
3.
ADB prepared the PCR in 2010 based on available data and concluded that, as
of 2007, achievement of the targeted health indicators2 had fallen short. However,
the status of these targets was not known at the time of the PCR. Since then,
the Indonesia Demographic and Health Survey (IDHS) of 2012 has been completed and
preliminary results have been published, and this helped the evaluation team to
ascertain the degree of achievement of targets.
4.
One of the key elements of project justification was that investment in public
health through the project would yield considerable economic benefits and substantial
cost savings.3 The economic analysis at project appraisal was based on disabilityadjusted life year and regional gross product data in each of the seven provinces, but it
was not updated in the PCR. Hence, it is not known to what extent the reported
benefitcost ratio in the range of 1.1 to 11.8 was realized. The PCR rated project
performance less efficient based on process indicators such as contract awards,
disbursements, and implementation delays.
5.
The rationale given in the PCR for preliminary assessment of project
sustainability (sustainable) was solely based on districts budget allocation for health.
While it states that health-care reforms were adopted in all districts medium-term
development plans, it does not say whether all or some reforms were implemented and
continued by the central and district governments. The PCR states that several
innovations were launched under the project (footnote 3, para. 16) but that it was not
clear how successful these innovations were. Hence the PCR suggested that an
1
2

ADB. 2010. Completion Report: Decentralized Health Services Project in Indonesia. Manila.
The indicators included maternal mortality ratio, infant mortality rate, under-5 mortality rate,
and life expectancy at birth. In addition, four outcome (purpose) targets included contact
rates, unmet need for family planning, financial sustainability of health facilities, and the
sustainability of local health safety-net programs.
ADB. 2000. Report and Recommendation of the President to the Board of Directors: Proposed

Loan and Technical Assistance Grant (financed from the Japan Special Fund) to Indonesia for
the Decentralized Health Services Project. Manila, paras. 114115.

PCR suggested
an evaluation of
the efficiency
and
sustainability of
the various local
health initiatives
initiatives

Decentralized Health Services Project in Indonesia


evaluation of the efficiency and sustainability of the various local health initiatives was
needed before replicating the initiatives. It is not known if such an evaluation was
made and if some or all innovations were replicated.
6.
The evaluation is based on a review of program documents, various country
studies, and the findings of an independent evaluation mission to Indonesia.4
The evaluation team undertook additional analysis to supplement data collected during
the mission and data provided by the Directorate of Nutrition and Maternal Child
Health.
7.
The evaluation report has been prepared 3 years after the PCR was issued and is
based on IED guidelines.5 It provides an in-depth analysis of relevant evaluation
parameters, including economic reevaluation. It focuses on project outcomes and
impacts as well as sustainability of benefits attained during the project period.

B.

The project
aimed to
improve the
health status

Expected Results and Project


Project Objectives

8.
The project aimed to improve the health status of the population in the project
areas. Its objectives were:
(i)
improved health and family planning services, and
(ii)
guaranteed access of the poor to essential health and family planning
services.
9.

The project activities were clustered into four outputs:


(i)
advocacy and capacity building;
(ii)
adapting health services to local needs through appropriate health
sector reforms;
(iii)
investments in health and family planning; and
(iv)
project management, including implementation, monitoring, and
evaluation.

10.

Technical assistance (TA) attached to the project6 envisaged six outputs:


(i)
health sector reforms identified,
(ii)
capacity for health system management and health service delivery
built,
(iii)
health plans developed,
(iv)
awareness of health priorities created among decision makers,
(v)
operations research capacity developed, and
(vi)
project management supported.

The mission team held discussions with staff at the Directorate of Nutrition and Maternal and
Child Health, and project staff of the Second Decentralized Health Services Project (DHS2);
made field visits to selected sites in North Sulawesi, Bali, Central Sulawesi, and Riau Islands;
held discussions with development partners such as the Australian Agency for International
Development (AusAID), the German Society for International Cooperation; United States
Agency for International Development (USAID), World Health Organization (WHO), and
collected inputs from health facility staff.
IED. 2006. Guidelines for Preparing Performance Evaluation Reports for Public Sector Projects.
ADB: Manila.
ADB. 2000. Technical Assistance to Indonesia for Decentralized Health Services. Manila
(TA 3448, approved on 26 May 2000 for $180,000).

Introduction

C.

Project Completion Report Assessment

11.
ADB approved the PCR in January 2010, and it rated the project successful.
The project was assessed as relevant, effective, less than efficient, and likely to be
sustainable. It was consistent with ADB's strategy and met most stated objectives.
When the project began, decentralization had not adequately clarified the devolution
of functions to different tiers of government, which made implementation of the
project, and local provision of health services, a challenge. This led to implementation
delays, and consequently the project was deemed less than efficient.
12.
Regarding the outputs, the project was effective and met many appraisal
targets for the physical infrastructure and training. Regarding the outcome, the project
was assessed as effective. It contributed to better health and family planning service
delivery in the project areas. It helped develop capacity for local health planning and
management, and bolstered the technical skills of doctors, nurses, village midwives,
and paramedical staff. The PCR notes that the institutional agenda could have been
more effective if the phasing or timing of policy reforms, resource constraints, and the
political will of stakeholders had been realistically assessed and factored in.
13.
The PCR also states that the TA produced several documents on local health
sector reforms and training modules for health planning and budgeting, which the PCR
deemed flexible and relevant. The project was originally designed to cover 50 districts
and 12 cities in 7 provinces, but as decentralization evolved the coverage was
expanded to 73 districts in 8 provinces. Of the 19 major covenants, 15 were complied
with, 2 were partly complied with, and 2 were complied with delays. Partial compliance
resulted from the late release of counterpart funds and the failure to recruit an
international project management firm. Late compliance resulted from the late fielding
of the midterm review mission and late submission of the government's PCR.

CHAPTER 2

Design and Implementation


14.
The chapter draws on review of relevant documents, discussions with key
informants, and data provided by the project. It discusses project formulation and
rationale behind the project and summarizes cost, financing and implementation
arrangements; procurement, construction and scheduling; and design changes. It also
highlights four key outputs achieved under the project and reviews consultant
performance and loan covenants associated with the project.

A.

Project
preparation
lacked detail

Formulation

15.
Overall, the evaluation noted that project preparation lacked detail. Local
governments developed preliminary investment programs to meet local health issues
and challenges. Due to the brevity of planning, it was agreed that the project would be
revised after district and provincial staff had received further training in health planning
and management. Consequently, implementation was done in two phases. The first
phase focused on training and capacity building, with consolidation of local health
sector development plans into more comprehensive sector plans. These plans were
revised by the executing agencythe Ministry of Health (MOH)and approved by ADB
following certain criteria defined in the loan documents. The plans were implemented
in the second phase of the project.
16.
Given the limited time for project preparation there were no detailed reports on
public health, medical equipment, and procurement needs in each province; and
assessments of training needs assessments for operational and management staff.
According to the PCR, the TA produced a number of documents on local health sector
reforms and training modules for health planning and budgeting, but this input was
provided only midway through implementation.

B.

The central
government
devolved some
of its healthhealthrelated
responsibilities
in 2001

Rationale

17.
At the time of project formulation, health service delivery in Indonesia was
highly centralized, with limited management capacity at the district and local levels.
As a result, health services were perceived to lack responsiveness to local issues.
The central government devolved some of its health-related responsibilities in 2001,
in line with laws governing regional autonomy and passed in 1999.7 At the time of
project design it was deemed important that access to essential health services for the
poor, vulnerable groups, and women be sustained during the transition because the
country was still recovering from the Asian financial crisis 19971998. The Social Safety
Net Program for Health (JPS-BK),8 which had been established with ADB support to
7
8

Law on Regional Autonomy (Law No. 22/99) and Law on Fiscal Balance (Law No. 25/99).
ADB. 1998. Report and Recommendation of the President to the Board of Directors: Proposed

Loans and Technical Assistance Grants (financed from the Japan Special Fund) to Indonesia for
the Social Protection Sector Development Program. Manila (Loans 1622-INO and 1623-INO),
approved on 9 July 1998 for $300 million (with 50.8% of the loan amount for the health
sector); and ADB. 1999. Report and Recommendation of the President to the Board of

Directors: Proposed Loans and Technical Assistance Grants (financed from the Asian Currency

Design and Implementation

protect vulnerable populations from the adverse impact of the crisis, also needed
financial replenishment.
18.
The project design team noted that a reformulation of the JPS-BK was required
considering the challenge of decentralization with better targeting and cost-effective
delivery of health services. Nutrition, reproductive health, and communicable disease
control remained key issues for the health sector, along with the introduction of
capacity to reform health-care financing and human resource development.
In addressing these issues, the government faced a considerable resources gap since
the JPS-BK was largely externally financed. In response to this need, the government
requested ADB's support in early 2000 to help implement decentralization in the health
sector. DHS was approved on 14 December 2000 and became effective on 25 June
2001.
19.
At appraisal, the government had given high priority to sustaining health
service delivery at a time of major upheaval in the architecture of the public sector
health system. The project was undertaken at a time when central government services
were being devolved and limited capacity for implementation was apparent at lower
government levels. At the midterm of implementation, the government accorded high
priority to improving womens health and project activities were reorganized to meet
this demand. The project provided support for implementing governments priority.
ADB had also accorded high priority to improving womens health in its 1999 Health
Policy, and so this project was in line with the priorities of both the government and
ADB. Indonesias health policy, Healthy Indonesia 2010, highlights the need to
support primary health care and boost expenditures for health.9
20.
In the Millennium Development Goal (MDG) progress report, the MDG targets
associated with maternal health in Indonesia required special attention, particularly in
relation to skilled attendance at child birth.10 Leading causes of maternal mortality
included high-risk pregnancies and unsafe abortion. Constraints such as inadequate
family planning, limited access to basic and emergency obstetric and neonatal care,
limited capacity of health professionals in some areas, and a weak referral system
hinder improvements in maternal health. These considerations make the project
relevant in both the current and past contexts of the need to improve womens health.

C.

Cost, Financing, and Executing Arrangements

21.
At appraisal, the project was estimated to cost $87.0 million, with ADB
providing a loan of $65.0 million equivalent (special drawing right [SDR] 50.5 million)
from its Special Funds resources. The government was to provide $22.0 million mainly
to finance taxes and duties, recurrent costs during implementation, and a portion of
the civil works, training, consultants, operations research, and project management.
By output, actual expenditures for advocacy and capacity building were 34% above
those estimated at appraisal. Actual project management expenditure was 40% higher
than that forecast at appraisal, largely because the project implementation period was
much longer than planned. Only the operations research and pilot-testing output

10

Crisis Support Facility) to Indonesia for the Health and Nutrition Sector Development Program.
Manila (Loans 1675-INO and 1976-INO), approved on 25 March 1999 for $300 million.
C. Rokx et al. 2010. New Insights into the Provision of Health Services in Indonesia: a Health
Work Force Study. Washington, DC: World Bank.
The National Development Planning Agency of Indonesia. 2011. Report on the Achievement of
the Millennium Development Goals, Indonesia 2011. Jakarta.

MDG targets
associated with
maternal health
in Indonesia
required special
attention
attention

Decentralized Health Services Project in Indonesia


(Adapting Health Services to Local Needs) incurred 17.4% less expenditure than
planned ($5.369 rather than $6.501 million).
22.
Disbursement by cost category reflects differences in output specific
implementation. Output 3 (Investments in Health and Family Planning Services) was
largely implemented as planned, so planned and actual disbursements for civil works
and equipment are similar. For example, $12.629 million was the planned expenditure
for civil works and $12.707 was disbursed. However, only 23% of the planned
expenditure on domestic and international consultants ($0.833 of $3.630) was utilized.
The implication of the shortfall is summarized in section G of this chapter.
23.
The PCR noted that as a result of the appreciation of SDR against the US dollar,
the dollar equivalent of SDR50.5 million increased from $65.0 million to about
$72.0 million. When loan cancellations of $12.6 million are considered, the
undisbursed loan balance was only $3.3 million. The actual expenditure stood at
$91.6 million, comprising $59.3 million in ADB financing and $32.3 in government
funding. Thus ADBs actual contribution was 64.7% of the project cost instead of the
originally envisaged 74.7%.

The provincial
focus of the
design was
inappropriate

24.
MOH established a central project management unit for managing
implementation of non-family-planning activities, along with provincial management
units in each of the project provinces. District implementation units were responsible
for monitoring project implementation in subdistricts and villages. Some of the key
informants told the evaluation team that the provincial focus of the design was
inappropriate given that the devolution of health system management to the districts
required district-specific project management capacity and training.11
25.
At appraisal, the World Bank (footnote 9) noted an intermediate role for the
provinces in technical support, disease surveillance, health promotion, procurement,
and personnel management.12 At that time, however, it was unclear how functions
would be distributed between central, provincial, and district authorities, even though
the Law on Regional Autonomy had decentralized most health-related responsibilities.
Consequently, ADB adopted a somewhat centralized project management structure for
its follow-on project, the Second Decentralized Health Services Project (DHS2),13 since
regional lack of professional and management skills was found to be the greatest
constraint on implementation.
26.
The family planning component was implemented by the National Population
and Family Planning Board (BKKBN). It established a central management unit and
secretariats in each of the project provinces. Institutional issuessuch as family
planning being integrated into local government offices, resulting in the loss of staff,
and the absence of a joint technical committee (with representatives from MOH and
BKKBN)hindered project planning and implementation.

11

12

13

The design of the follow-on project, DHS2 (footnote 13), thought to reflect the lesson of not
having this focus and established greater district management control over implementation.
World Bank. 2008. Provincial Health Project in Indonesia. Washington, DC.
Available: http://documents. worldbank.org/curated/en/2008/06/9736257/indonesia-provincialhealth-project
ADB. 2003. Report and Recommendation of the President to the Board of Directors: Proposed
Loans to Indonesia for the Second Decentralized Health Services Project. Manila.

Design and Implementation

D.

Procurement, Construction, and Scheduling

27.
The project suffered considerable and repeated delays in civil works and in
procurement of medical equipment and supplies. The key reason was the late release of
counterpart funds, which narrowed the effective implementation window to only
6 months per year. Once funds were available progress was considered satisfactory.
However, the continued delays resulted in two project extensions. Liquidity issues after
the Asian financial crisis were noted as a possible cause in the evaluation teams
discussions with stakeholders.14

The project
suffered
considerable
and repeated
delays

28.
Civil works, which mostly involved either rehabilitation or rebuilding on the
same land, were mostly awarded through competitive bidding. The PCR reported the
performance of the civil works contractors as generally satisfactory. The quality of
facilities inspected by the evaluation team was deemed good, and minor needs for
repair (in roofing) and some drainage issues were to be expected given the age of
infrastructure. Several key informants told the evaluation team that the lengthy bidding
process under ADB guidelines, including letters of no objection, caused unnecessary
delays to project implementation.
29.
Contracts with an estimated value below a ceiling amount agreed during loan
negotiations did not require a letter of no objection from ADB, but that ceiling had
been set at a low level of $50,000. Taking advantage of decentralization, some
procurement and civil works contracts were divided into smaller, district-specific
packages that often stayed below the ceiling, but because they had originally
(at appraisal) been estimated to be above this threshold there was confusion about the
proper procurement method. Better specification of civil works and equipment at
design, and during implementation, might have helped overcome some of these
problems, as would have TA on ADB processes.
30.
The evaluation team found that all equipment procurement had been
completed and inspection of facilities showed much of the equipment to be in good
order. Some medical equipment was coming to the end of its useful life. Most health
departments staff indicated that local government finances were sufficient to replace
basic equipment. Some noted the need for more sophisticated scanning equipment,
but these could not be supported using local resources. The evaluation team observed
that bed capacity was reaching the limitbed occupancy rates exceeded 80%, in part
because of expanded health insurance coverage. Hospital managers indicated that
access to government finance was limited to covering the costs of capital equipment
and precluded extensive infrastructure development.
31.
The project was to be implemented over 5 years, with physical completion
scheduled for 30 September 2006. It was physically completed on 30 June 2009, after
two extensions. The major reasons for delay have already been discussed. Much of the
delay occurred before the midterm review. In February 2004 (55% elapsed loan period),
project implementation progress was 46%, cumulative contract awards were at 21%,
and disbursements had reached 16%. At the time of the final loan mission, the project
had achieved physical implementation progress of 89% against an elapsed loan period
of 98.6%. Civil works were completed in 2006 and all goods had been delivered.
By 18 November 2008, all proposed contract packages had been awarded and
completed.

14

The DHS2 has not encountered these issues, so the problem appears to have been rectified.

The evaluation
team observed
that bed
capacity was
reaching the
limit

8
Lengthy
procurement
processes
were
compounded
by the lack of
capacity
capacity

Decentralized Health Services Project in Indonesia


32.
Issues associated with delays in counterpart funding and lengthy procurement
processes were compounded by the lack of capacity to develop district implementation
plans during the startup phase of the project. The lack of capacity and consequent time
taken to formulate local investment plans was highlighted during the evaluation teams
field visits, and suggests that the phasing and scope of the project was too optimistic
given the uncertainty surrounding decentralization in Indonesia. The delay in mobilizing
TA to help develop local planning capacity also undermined implementation efficiency,
and given the size and geographic scope of the project, the original TA value was
insufficient.

E.

Design Changes

33.
At the request of the government, ADB provided associated TA to assess
project achievements and gaps.15. The analysis concluded that, among other things,
the outcomes supported by the project were not on course to meet the 2010 targets
set for the maternal mortality ratio (MMR) and the under-5 mortality rate (U5MR).16
Under the TA, at the request of the project manager, the consulting team formulated a
comprehensive grand strategy to refocus on maternal and newborn child health
(MNCH). The government approved the strategy, which had five pillars:
(i)
Improve the supply side of essential MNCH services in terms of access,
quantity, and quality.
(ii)
Stimulate community demand for MNCH and family planning services.
(iii)
Improve cooperation between the public sector and private and
nongovernment organizations (NGOs).
(iv)
Increase the capacity of district health offices to manage MNCH.
(v)
Develop local political commitment and funding for health and family
planning from local government and Regional House of Representatives
(Dewan Perwakilan Rakyat Daerah).

The new focus 34.


According to key informants, the new focus paid more attention to MNCH and
attributable to
paid more less attention to family planning and nutrition. This was partly
reorganization within MOH and changes in project responsibilities.17
attention to
MNCH F.
Outputs
35.
The project had four outputs. Most were delivered as planned, although
expenditures for output 2, which aimed to pilot local reform initiatives and to increase
service efficiency, were 17.5% less than at appraisal ($5.369 million compared with
$6.501 million). For all other outputs, planned and actual expenditures were close.
1.

Output 1: Advocacy and Capacity Building

36.
Output 1 primarily supported the training of staff in health system
management and to a lesser degree in clinical skills, mainly in the area of MNCH.
The first phase of the project included advocacy, social mobilization, and workshops to
15

16

17

ADB. 2000. Technical Assistance to Indonesia for Support for Health Sector and Policy Reform.
Manila (TA 3579), approved on 14 December 2000 for $1 million, associated with Loan 1810INO (footnote 3).
Ministry of Health. 2008. Completion Report: Decentralized Health Services Project in
Indonesia. Jakarta (p. 10).
At the time of the midterm review in 2005, the project was under the Directorate General of
Community Health, a change from the earlier arrangement under the Secretary General
(Project).

Design and Implementation

assist implementation, and involved 31,000 participants (52.3% male and 47.7%
female). The project developed collaboration with local universities and teaching
institutions with expertise in public health during this phase. Advocacy was used by the
district health offices to receive support from local government and political
commitment from local parliament, and from various local community leaders.
37.
Various courses were offered, including short ones (less than 1 week). Diploma
courses included midwifery, nursing, and nutrition, while undergraduate courses
covered health management, health promotion, statistics, and nursing. Master courses
were provided both in Indonesia and abroad in field epidemiology, reproductive health,
health promotion, public health, and health economics. PhD courses on health
management, and policy and nutrition, and specialized training for cardiologists and
pediatricians were also supported. In all, 3,032 staff benefited from fellowships
(Table 1) and nearly 56% of the participants were women. The proportion of women
participating in bachelor's degree, master's degree, and specialist courses was
unfavorable, howeverthe share of female staff trained is well below 50%.
Table 1: Number of Fellowships under the Project by Staff Level

Year
2001
20012005
2006
20062008
Total
Female %

D-III
1,231
50
1,281
75.5

D-IV
62
6
68
92.6

S1
655
134
789
43.5

Staff Level
S2
Spec
575
21
219
2
794
23
38.5
34.8

S3
8
0
8
12.5

S2 (os)
32
6
38
50.0

SC
36
18
54
46.3

Total
2,599
433
3,032
55.6

D-III = diploma 3, D-IV = diploma 4, S1 = bachelor's degree, S2 = master's degree, S2 (os) = short course
(overseas), S3 =doctoral degree , SC =short course (in-country) , Spec = specialist
Note:
Based on Law No. 20/2003 on the National Education System, education in Indonesia consist of
basic education, middle education, and higher education. Higher education encompasses: diploma,
sarjana, magister, specialist, and doctorate. Diploma (D) is shorter and more practice-oriented than
sarjana/bachelor (S). Diploma 4 (D-IV) is equal to the bachelor's degree (S1). After finishing D-IV or
S1, a student can move to S2 (magister) or specialist (Spec) education (to become a doctor, for
instance). Diploma levels are: diploma 1 (2 semesters), diploma 2 (4 semesters), diploma 3
(6 semesters), diploma 4 (8 semesters).
Source: Independent Evaluation Departmentdata collected by the evaluation team during field visits.

38.
A consortium was established to coordinate the training program. Its members
included BKKBN and the Ministry of Health and Social Welfare, university staff
(national, regional, and local universities), and representatives of professional bodies
(doctors and nurses) and NGOs. The main roles covered provision of technical support
to formulate and implement human resource development activities and action
programs; mobilization of necessary resources for training units; and evaluation of the
impact of human resource development. The consortium established a list of training
needs, then started developing corresponding modules for both the trainers and the
trainees. To date, eight modules have been developed.
39.
Management training involved 21,000 participants (57% female and
43% male). It aimed to improve the capability of health staff in the management of
decentralized health services, decentralized health facilities, and priority health
programs. The key informants cited capacity development as one of the major benefits
of the project. Many respondents reported that local health resources could not
support such training, although it had been instrumental in the career development of
many current senior health officials. Key benefits listed by interviewees were: (i) better
planning, (ii) module development for local government training, (iii) better
organizational planning, (iv) ability to budget, (v) capacity to mobilize the community,

Many
respondents
reported that
local health
resources could
not support
such training

10

Decentralized Health Services Project in Indonesia


(vi) enhanced management of public health center facilities, (vii) problem solving, and
(viii) networking.
40.
Technical needs-based training covered lifesaving skills for midwives, basic
emergency obstetric or neonatal care, normal delivery care, and comprehensive
emergency obstetric and neonatal care. Training focused on developing the skills of
midwives at health centers and village delivery centers (polindes) that provide ante- and
postnatal care and assist with deliveries.

Midwives
indicated that
the training
had been of
value

41.
The PCR noted that the project trained about 40,000 midwives and nurses,
most of them district based, on lifesaving skills over 2 years. Midwives interviewed by
the evaluation team indicated that the training had been of value and pointed to a
need for follow-on training to refresh their skills, and to train new cohorts of midwives.
Stated benefits from the training program were: high-risk pregnancy identification,
improved referral capacity, antenatal care, emergency obstetrics, post-delivery care,
and access to MNCH kits. However, the local budget was viewed as insufficient to meet
the demand for MNCH training.
42.
According to the PCR and based on discussion with BKKBN senior staff,
capacity building in family planning covered: (i) training and fellowships for
reproductive health and family planning; (ii) civil works to renovate warehouses,
training centers, intensive and critical care facilities, family planning clinics, and various
offices; (iii) advocacy and information, education, and communication concerning
family planning; (iv) equipment for family planning counseling centers, advocacy, and
communication; and (v) operations research aimed at improving contraceptive use by
men and young adults. A total of 153 district staff benefited from in-country fellowship
programs and degree courses (D3 to S-1), especially field workers in family planning.
They also confirmed that refocusing the project largely on MNCH meant that family
planning activities were limited.
2.

Strong
government
commitment
and local
community
ownership
ownership
played a
significant role
in sustaining
the reforms

Output 2: Health Sector Reforms

43.
The second output aimed to pilot reforms to increase the efficiency and quality
of health services. Activities followed the two-step approach of the projectconsulting
services initially supported local health sector analysis and, later, selection of locally
suitable reform proposals (e.g., health-care financing, improving public financial
management, streamlining human resources, and publicprivate partnerships) to be
included in comprehensive health sector plans. Priority was to be given to meeting the
needs of the poor and vulnerable groups. ADB reviews noted that the TA helped
highlight key issues for local health sector reform and documented various case studies,
while the PCR stated that the pilot covered about 40% of the project districts, including
a partnership between traditional birth attendants and midwives, community-based
health financing, and a total quality management model. Examples of reforms to boost
the quality of health services and increase health-financing schemes to improve the use
of health services are summarized in Table 2. Informants noted that strong government
commitment and local community ownership played a significant role in sustaining the
reforms introduced under the project.

Design and Implementation

11

Table 2: Health Sector Reforms Introduced and Their Status


1.

Bakesra (Balaikesehatanrakyat) Southeast Sulawesi Rural and Remote Health Post

The initiative aimed to improve access to health services and community empowerment, particularly for rural and remote areas.
It was piloted in Kolaka District of Southeast Sulawesi covering 18 subdistricts and 30 villages. It led to increased coverage of
antenatal care, immunization, better nutrition, and health promotion in the intervention areas. Findings indicated that a healthy
lifestyle is a good approach to community empowerment and Bakesra has the potential to develop into an insurance program like
Akesra. The initiative inspired the development of national Desa Siaga. However, Bakesra was discontinued due to the shift in
attention to JamkesmasJamkesda, Jampersal, and BOK.
2. Jaminan Kesehatan Jembrana 20032003-2007
The initiative was piloted in Jembrana District, Bali to investigate the potential of universal health insurance with an aim to
improve equity and quality of health services. It covered 6 subdistricts and 51 villages. It assisted renovation of buildings,
provision of medical and nonmedical equipment, as well as fellowships for staff supported by Bhupati. The results have been
encouraging. About 78% of the districts population is covered under the scheme, and residents have health cards that entitle
them to get free primary and secondary (hospital) health care. The pilot has introduced integrated health services offered by
private providers. It has been successful in unifying health budgets and providing health services down to the village level.
Regulations on tariffs, standard operating procedures for services, pooling of resources, and promotion of public-private
partnerships in health service delivery were developed as part of the pilot. The program is now adopted across all districts of Bali
Province.
3. Provide Family Doctor In Riau Islands Province
Riau Islands Province is a geographically difficult area to serve because the population is spread over many islands. This initiative
was launched to overcome shortage of health facilities and staff, including doctors and nurses, in the archipelago with an aim to
improve equity in access to quality health services. It was piloted in Tanjung Balai Karimun District covering 6 subdistricts and
65 villages. The initiative started in 2006 with only two medical doctors and had expanded to 35 doctors by 2008. The program
allows home visits by contracted doctors and nurses to families in remote areas, and on average one doctor is assigned to
118 families in a month. The experience has been promising. As a result of this initiative, (i) families have access to comprehensive
health services, (ii) local people have demonstrated change in their behavior and have adopted healthy lifestyles, (iii) school health
programs are running effectively, and (iv) delivery of services has been more cost-effective than using motorboats as mobile
health centers. The program has been scaled up and adopted in all districts of the province with remote communities. At the time
of the evaluation teams field visits, 115 contracted doctors were serving remote areas. District health offices provide buildings
and medicines. The success of the initiative was noted as being driven by the governors strong commitment.
4. Midwives - Trained Birth Attendant Partnership
The pilot program was launched in three subdistricts of Siak District (Riau Province) in 2004 with an aim to enhance community
empowerment and focused particularly on improving health services and reforming the referral system. Under this initiative the
local government provided Rp100,000 per month as an incentive for trained birth attendants (TBAs) to bring pregnant women to
midwives. It offered better access to child delivery with the support of trained health workers, particularly for women from
economically disadvantaged groups, which also helped reduce maternal and neonatal mortality rates. Better recording and
reporting of MNCH activities and outcomes were noted as another result of the pilot. The initiative is ongoing, but incentives
given to TBAs have been discontinued owing to lack of funds. The impact of this has not yet been assessed.
A similar pilot was launched in Minahasa District, North Sulawesi with the same objective, covering all villages and all subdistricts.
The program has helped improve antenatal care, delivery by trained health workers, and an integrated local budget for MNCH.
Skillful midwives and TBAs made the program sustainable, and other districts have replicated the model. The partnership initially
established still exists.
5. Manado Clean and Green City
With project support, the initiative was introduced in Manado City, and in 9 subdistricts and 30 villages. Its focus was on
changing human behavior through community empowerment. The initiative has been institutionalized and both the public and
private sectors are actively participating. As a result, noticeable behavior change has been observed in most of the communities.
The national committee has established awards for communities demonstrating lower diarrhea incidence. The program continues
to date, largely thanks to government commitment, and has been replicated by other districts and Bitung City.
MNCH = maternal, newborn and child health
Source: Ministry of Health, Indonesia.

12

Decentralized Health Services Project in Indonesia

3.

Output 3: Health Services Infrastructure

44.
Output 3 had the objective of building and renovating health service
infrastructure and providing medical and nonmedical equipment. Civil works included
the construction or rehabilitation of hospitals, health centers (puskesmas), and village
health posts (posyandus).18 Given the project's objective of improving access to health
services for the poor, lower-level facilities were to be the target of most construction
activities. This focus varied a great deal across provinces (Figure 1): North Sulawesi had
the largest proportion of overall construction resources earmarked for puskesmas and
posyandus, Bali had the lowest. In Bali, construction expenditure was dominated by the
rehabilitation of Policlinic Amlapura Hospital, Indera Hospital, Tabanan Hospital and
Klungkung Hospital. Nanggroe Aceh Darussalam (NAD) had large hospital expenditures
but also spent large sums on construction of district and provincial health department
buildings. Following the 2004 tsunami, Indonesia received an influx of funds from
other sources and as a result a large amount of health center investment was
reprogrammed by the project management.
Figure 1: Proportion of Civil Works Expenditure
by Facility in Project Provinces

Other

Health Centre

Hospital

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

Total

NAD

Riau

Bengkulu

Southeast Sulawesi

Central Sulawesi

North Sulawesi

Bali

Kepulauan Riau

NAD = Nanggroe Aceh Darussalam.


Source: Independent Evaluation Mission data.

45.
The IED evaluation team visited a selection of health facilities and talked with
relevant staff. It found that all but one of the health facilities supported by ADB were in
good order, and district health officials were positive about the design and relevance of

18

The poor most often use puskesmas and posyandus because access is easier and they are
cheaper than hospitals.

Design and Implementation

13

construction undertaken under the project.19 Medical and nonmedical equipment


(including vehicles) accounted for $20.8 million (23% of total project cost).
The provision of motorcycles to the midwife coordinators strengthened outreach
activities. Developing these facilities helped promote utilization of health insurance.
BKKBN investments included (i) civil works such as renovation of school rooms for
adolescent reproductive health, information and counseling centers, and contraceptive
storage room at a cost of $1.1 million, and for village family planning clinics; and
(ii) information, education, and communication equipment and vehicles at a cost of
$ 2.4 million.

Developing
these facilities
helped promote
utilization of
health insurance

46.
It should be noted, however, that the PCR cites reports by the National Audit
Authority according to which the provision of equipment under the DHS was generally
not satisfactory because the delivery of goods (e.g., ambulance boats and midwifery
equipment kits) was incomplete, delayed, and/or not in compliance with specifications.
The audit also found that civil works did not conform to the original design, and some
were left incomplete. The evaluation team did not have enough time to verify these
audit findings, but MOH did not contest them.

The audit also


found that civil
works did not
conform to the
original design

4.

Output 4: Project Management

47.
Under the output, 2,580 staff received training in project management and
implementation. The actual expenditure exceeded appraisal estimates because of the
extended implementation period and the need to support central, provincial, and
district implementation committees for longer than expected. All planned activities
appear to have been implemented, although monitoring and evaluation (M&E) was less
than satisfactory. This issue is examined in the assessment of overall performance
(chapter 3).

G.

Consultants

48.
The project provided for 22 person-months of international consulting inputs
on policy and health system strategy, health service organization, and information
technology. Only 14 person-months were used during implementation. No information
technology input was used, but more than the planned amount of policy and strategy
inputs was used at later project stages.
49.
As for national consultants, the project used only inputs related to human
resources. It did not use others planned on health policy analysis, systems
development, hospital infrastructure and equipment, mother and child health, human
resource management, family planning and reproductive health management, gender
and health, and food and drug control.
50.
Feedback from the evaluation teams field visits indicates that the consultants
performance was satisfactory. Input by international policy consultants was reduced at
midterm because the project deemed it not relevant to Indonesias context. The PCR
notes that the project followed ADB's Guidelines on the Use of Consultants to recruit
consultants. The project recruited consultants on an individual basis because the
executing agency preferred to limit the number of international consultants. While
19

The Central Sulawesi provincial hospital was relocated to an area less prone to flooding.
The old building was in disuse at the time of the evaluation teams visit and had been
vandalizedbroken windows, torn roofing, and some items missing. The hospital had received
project support for renovations, which is deemed to be a sunk cost.

14

Decentralized Health Services Project in Indonesia


recruitment of individual consultants introduced flexibility, it made it harder for project
management to coordinate their inputs and outputs.

M&E suffered
from the
absence of
baseline
information
information

51.
The planned engagement of local consulting firms or NGOs to conduct social
and environmental impact assessments did not receive adequate attention.
M&E suffered from the absence of baseline information against which progress could
be measured. Furthermore, the health management information system introduced
under the project is deemed inadequate.

H.

Loan Covenants

52.
The PCR reported that compliance with major loan covenants was generally
satisfactory. Audited financial statements and quarterly progress reports were
submitted on time. Of the 19 major covenants, 15 were complied with, 2 were partly
complied with, and 2 were complied with late. Late compliance was associated with
the government's PCR and delays in the conduct of the midterm review. Partial
compliance reflected late disbursement of government counterpart funds and not
using an international firm for project management.

CHAPTER 3

Performance Assessment
53.
The chapter provides an overall assessment of the project performance and
discusses each of the four key performance assessment parameters. These include
relevance, effectiveness, efficiency, and sustainability.

A.

Overall Assessment

54.
The evaluation rates the overall project performance successful, which is
consistent with the PCR rating. The project helped improve the delivery of health
services and, to a limited extent, family planning services by refurbishing and/or
expanding local, district, and provincial health service facilities in eight provinces;
boosting management and technical capacity of health service staff; introducing
subsidies for traditional birth attendants, a model that has been expanded to other
parts of Indonesia; and implementing a health insurance scheme for the poor segment
of the population. Through capacity development initiatives, midwives could acquire
competency in obstetric and neonatal care. The project also played an important role in
increasing the share of health in district budget allocations.
55.
The project outcomes and impacts are yet to emerge clearly. The evaluation did
not find significant differences in key health indicators between the provinces
supported by the project and those not supported. Nevertheless, improvements are
noted in maternal and infant mortality rates in most of the provinces. Other provinces
may have had similar health interventions funded by other development partners
and/or governments internal resources. In addition, this may be due to an overall
national commitment to prioritizing support for better health outcomes. Table 3
summarizes the project performance evaluation ratings, followed by detailed
discussion. Appendix 1 outlines the achievements against the targets set in the design
and monitoring framework.
Table 3: Assessment of Overall Performance
Criterion
Relevance
Effectiveness
Efficiency
Sustainability
Overall
Overall rating

Weigh
Weight
(%)
25
25
25
25

Ass
Assessme
essment
Relevant
Effective
Less than Efficient
Likely Sustainable
Successful

Ratin
Rating
Value
2
2
1
2

Weigh
Weighted
Ratin
Rating
0.50
0.50
0.25
0.50
1.75

Source: Independent Evaluation Departmentevaluation team assessment based on Appendix 2.

B.

Relevance

56.
The evaluation rates the project relevant at design, during implementation,
at completion, and at the time of evaluation. In 1999 and compared with other Asian
countries, Indonesia had a high MMR of 420 maternal deaths per 100,000 live births,
and an infant mortality rate of 43 deaths per 1,000 live births. Decentralization

The evaluation
rates the overall
project
performance
successful

16

Decentralized Health Services Project in Indonesia


threatened to slow any improvement in, or even worsen, these health indicators
because at the time of this policy change districts and communities had limited
capacity and institutional processes.
57.
Development of human resources through training, workshops, and higher
studies accounted for about 29% of project expenditure. The allocation is deemed
justified given the need for provincial and district capacity to improve health service
delivery, plan infrastructure investments, and purchase medical and nonmedical
equipment. The project made provisions for piloting new approaches to health-care
reform that would improve access and quality of health services, particularly for
geographically and economically disadvantaged population segments.

A poverty and
social
assessment
was not
accomplished

58.
The project documents did not explicitly state a rationale for resource allocation
among central, provincial, and district levels. Allocations should have been based on
the extent to which they supported particular functions under a decentralized model.
Districts put forward proposals for funding of civil works, training, and equipment
within each project province. Access to essential health services for the poor, vulnerable
groups, and women was to be the overarching priority in any proposal for project
support. A poverty and social assessment was to be prepared on how the various
strategies and activities proposed for project support met the needs of the poor,
vulnerable groups, and women. This task was not accomplished.
59.
The project formulated guidelines for the preparation of a comprehensive
health sector plan, and for central, provincial, or district proposals for project financing.
After translation into Bahasa Indonesia the guidelines were distributed to all project
provinces and districts. One international consultant and two domestic consultants
visited all project provinces to assist local teams in preparing their plans. Ideally,
hospital support should have focused on level-three inpatients to ensure that poor
people benefited from the project.
60.
The project aimed to improve locally managed health care delivery, support
health sector reform, and strengthen health service infrastructure in the districts. The
inclusion of community-led planning reflected lessons from earlier projects. As noted in
the PCR, experiences made with the 1998 Social Protection Sector Development
Program and the 1999 Health and Nutrition Sector Development Program (footnote 8)
suggested a need to involve provinces or districts in the planning process to factor in
their particular needs.

CommunityCommunityled planning
results in
better health
outcomes

61.
While the evaluation concurs with the notion that community-led planning
results in better health outcomes, the project would have benefited from more detailed
situation and problem analysis, risk assessment, and consultation with different
stakeholders during its formulation. Such an approach might have ensured better
understanding of the needs and challenges of public health planning during
decentralization. That the project design anticipated devolution at the provincial rather
than the district level proved a key limitation. A more detailed approach might have led
to a more realistic assessment of constraints to achieving the specified targets, and the
formulation of strategies to overcome these hurdles. For example, the project suffered
from late disbursement of government funds, yet a strategy was not considered to
mitigate this riskdespite being highlighted in the appraisal design and monitoring
framework.
62.
The evaluation found baseline data on the targeted population and public
health centers to be insufficient. Information on catchment population for each health

Performance Assessment

17

facility, available number of health workers, inventory of equipment, and other


supplieswas lacking. Collecting available information and conducting needs and gaps
analysis as part of a comprehensive analysis of the health sector and its services and
facilities would have improved early project performance. The midterm review of the
project helped in a greater focus on maternal and child health. As a result, project fund
disbursement improved.
63.
The project design was ambitious in its scope and geographical coverage.
It originally included all districts in seven provinces, covering more than 19 million
people in 50 districts and 12 cities (in Bali, Bengkulu, Riau, Central Sulawesi, NAD,
North Sulawesi, and Southeast Sulawesi). The number of provinces increased from
seven to eight and the number of districts from 50 to 73 after the inclusion of
Gorontalo Province in 2005. The increase in the number of districts was partly due to
the division of existing districts into smaller entities. Based on an estimated district
health expenditure of $38 per person, project expenditure averaged some 1%2% of
total district spending over the implementation period. Given this low level it is difficult
to ascertain why target outcome indicators were set so ambitiously.
64.
Health infrastructure and local capacities in these provinces were fragmented at
the start of the project. While these conditions indicated the need for the project, they
also highlighted the potential constraints on achieving project objectives, especially
when activities and resources were thinly spread over a large area. The project was also
overoptimistic given the complexity and uncertainty of the legal and economic
dimensions of decentralization. For example, many staff at the lower levels of the
health system had no experience in implementing donor projects. The large project
area also made it difficult to provide the necessary capacity development training to
meet the implementation schedule. The lack of early consultant input to help formulate
a comprehensive capacity development strategy and set priorities for tackling such a
complex project only exacerbated the problem.
65.
These realities were not embedded in the project design. It would have been
more desirable to implement the project across fewer provinces and a more compact
area. The selection of provinces was reportedly based on an outcome of donor
coordination at the time of project preparation, along with the criteria of willingness,
commitment, and readiness of local authorities to participate in the project. Both rich
and poor provinces were eligible. Given the pro-poor objective of the project, it would
have been of greater value to select provinces with higher poverty incidence and
greater health needsas defined by indicators such as mortality rates and life
expectancywhile still factoring in coordination with other stakeholders, and health
infrastructure and human resource capacity. Clustering these provinces into a more
manageable area would have made implementation support easier and capacity
development efforts less costly. The original intention of the project (at the time of TA
fact finding) was to limit the number of provinces to four. It is unclear why the
geographic scope increased so much, although the government apparently requested
(during pre-loan fact finding) that the project cover as many provinces as possible.
66.
Given the institutional uncertainties of decentralization, a project loan coupled
with TA was the proper lending modality for this project. ADBs Operations Manual
notes that implementation experience with past policy-based loans, comprehensive
sector analyses that identify structural constraints to sector development, and sound
policy and institutional settings are prerequisites for making financing commitments

The project
design was
ambitious in its
scope and
geographical
coverage

18

Decentralized Health Services Project in Indonesia


associated with program lending.20 These conditions or analyses were not evident at
appraisal. Moreover, the report and recommendation of the President went to some
length to detail the uncertainties surrounding decentralization, yet adequate TA to
support the project loan, along with realistic geographic coverage, was not factored
into the design.
67.
The evaluation team noted during the field visits that lower levels of the health
system, most frequently used by the poor, were supported by the project and continue
to operate. Pro-poor fund allocations did not seem to follow a geographic priority.
Many of the districts visited by the evaluation team were close to urban centers where
poverty rates are not so pronounced.

The districts
with higher
poverty
headcount, or
greater
percentage of
people living
below the
poverty line,
received the
lowest share
of civil works
funding

68.
Plotting district-specific civil works expenditure against poverty headcount in
each district (or city) illustrates a pro-rich allocation of resources (Figure 2). The districts
with higher poverty headcount, or greater percentage of people living below the
poverty line, received the lowest share of civil works funding.21 Other ADB health sector
support projects, which include the Health Care in the Central South Coast Project in
Viet Nam,22 did adopt geographic allocation formulas based on poverty headcount to
distribute the envelope of loan resources for civil works and equipment to provinces in
a pro-poor fashion. Although planning should be community-led, some guidance such
as formulas for broad allocation of funds could have helped align resource allocation
with project objectives.
Figure 2: Value of District Civil Works Expenditure vs. District Poverty Headcount
(% below poverty line)

Source: Independent Evaluation Mission data.


20
21
22

ADB. 2013. Operations Manual Section D4/BP. Manila, issued on 1 April 2013.
Correlation coefficient of 0.22.
ADB. 2008. Report and Recommendation of the President to the Board of Directors: Proposed
Loan to Viet Nam for the Health Care in the South Central Coast Region Project. Manila
(Loan 2468-VIE, approved on 7 November 2008 for $72 million).

Performance Assessment

69.
Despite the above-mentioned shortcomings in the project design, district
health services were encouraged to identify health priorities and key areas of healthcare reform. Health-care reforms were piloted in many instancessuch as the insurance
scheme in Bali and subsidies for traditional birth attendants. Much of the construction
centered on lower-level health facilities that are primarily used by the poor. Health
facility use by people in lower wealth quintiles is particularly low in Indonesia, so the
targeting of infrastructure to improve health outcomes for this group is considered still
relevant at evaluation.

C.

1.

Improved health and family planning services

71.
An aim of the project was to increase contact rates by 25%. Contact rates were
selected as an indicator because decentralization was seen as a major risk to the
provision of social sector services and social infrastructure.23 ADBs health sector policy
highlights access to basic health services as a core objective.24 Contact rates increased
in NAD, Riau, North Sulawesi and Central Sulawesi (Table 4). Outpatient contact rates25
increased in all project provinces except Southeast Sulawesi and Bali. Rates have
decreased or stabilized in these two provinces, although a trend in rate is not evident.
In Bali, 53.1% had an outpatient contact in 2000, which increased to 79.8% in 2003,
then decreased to 57.71% in 2006. The rate again increased in 2010, to 75.02%.
Although the trend is upward for most other provinces, there is large variation in yearto-year contact rates.

24
25

Much of the
construction
centered on
lowerlower-level
health facilities

Effectiveness

70.
The project's performance is rated effective in achieving outputs and outcomes,
confirming the PCR rating. The PCR noted that facility upgrades and staff training
helped improve the quality of services, especially patient services that are largely used
by the poor. For example, more births are now attended by skilled personnel. The PCR
also concluded, however, that both before and during project implementation external
factors might have equally contributed. The government launched new financing
programs such as Jamkesmas to support health service utilization by the poor,
Jampersal for maternal health, and block resources for the operation of lower-level
health facilities. Social Security Law No. 40/2004 was enacted with the objective of
universal social health insurance coverage. Finally, the growing involvement of the
private sector in Indonesian health care also influenced the use of health services. In the
absence of detailed M&E information it is difficult to quantify the role of project
support and other forms of health financing in these developments.

23

19

IED. 2010. Special Evaluation Study: Asian Development Bank Support for Decentralization in
Indonesia. Manila.
ADB. 1999. Policy for the Health Sector. Manila.
Number of persons experiencing illness who contacted a health care provider, divided by the
number of persons complaining of illness.

More births are


now attended
attended
by skilled
personnel

20

Decentralized Health Services Project in Indonesia


Table 4: Outpatient and Family Planning Contact Rates in Project Provinces

Province
NAD
Riau
Bengkulu
Riau Islands
North Sulawesi
Central Sulawesi
Southeast Sulawesi
Bali

Outpatient Contact Rate


(%)
2003
2006
2010
32.66
47.63
60.40
32.85
40.67
52.87
33.63
33.04
36.34
n.a.
40.07
52.09
10.50
43.54
48.33
31.38
37.70
56.55
87.18
30.07
78.41
79.76
57.71
75.02

Family Planning Contact Rate


(%)
2003
2006
2010
52.17
68.23
75.88
45.39
76.35
65.95
79.77
86.76
89.79
n.a.
78.66
65.06
70.67
78.93
84.46
n.a.
70.68
77.50
40.67
85.68
75.89
72.13
88.83
85.67

NAD = Nanggroe Aceh Darussalam, n.a = not available.


Source: Independent Evaluation Departmentdata collected by the project evaluation team.

72.
Much of the project investment focused on developing health centers.
Outpatient contact information captures patient presentation at health centers,
hospitals, and other government-run facilities. Contact rates have increased by 13.8%
to 17.4% across all project provinces between 2006 and 2010. This represents a 26%
increase, which is just above the project target of 25%.

Contact rates
for health
centers in
project
provinces
compared
with nonnonproject
provinces are
not
significantly
different

73.
When contact rates for health centers in project provinces are compared with
non-project provinces, they are not significantly different. Increases of this order across
the country suggest that factors such as health insurance expansion may be the main
driving force, rather than infrastructure improvement under the project. Recent sector
analysis indicates that primary care is still hindered by complex financing and reporting
arrangements, lengthy budget approval processes, and centralized human resource
management.26 The review team received positive feedback on the role of the project in
tackling some of these management constraints and in improving access to health
services. But to quantify the difference the project made on decentralized health
services, it is necessary to compare facility performance and client outcomes in project
and non-project districts. While project performance monitoring systems were included
in the project design, the evaluation team did not find a basis for detailed comparisons
and for an assessment of impacts. The evaluation team was made aware that data
collection in the public sector is sometimes hampered by poor supervision.
74.
ADB review missions noted that several local governments decided to charge
user fees for health services, but none of them had a pricing policy or reimbursement
regulations in place. They also made no systematic evaluation of the effect of their
decision on utilization by the poor and the vulnerable. This policy could have lowered
facility contact rates, and ideally the loan agreement should have set conditions to
prevent such a negative impact.
75.
The project included a target that at least 90% of deliveries be assisted by
trained birth attendants. Only Bali achieved this in 2005 (92%), all other project
provinces fell short.
76.
Family planning contact rates increased until 2006 but fell in 2010 in four of
the eight project provinces (Bali, Riau, Riau Islands, and Southeast Sulawesi), while only
marginal improvements are seen in the other four provinces (Table 4). Indonesias
family planning service at the time of devolution was a very centrally managed vertical
26

AusAID. 2011. Australia Indonesia Partnership for Health Systems Strengthening 20112016.
Program design document. Canberra.

Performance Assessment
program integrated into population offices within local governments. The budget
provided to family planning services decreased substantially. Consequently, several
family planning indicators are stalling and even falling in some provinces.
77.
While contact rates are increasing, contraceptive prevalence rates are not.
The prevalence of contraceptive use among never-married women and currently
married women aged 15-49 has been collected as part of the DHS. This indicator is
used for measuring the success of family planning programs. The data suggests that
the contraceptive prevalence rate has not improved since 2003 and has remained just
under 60%. The rates in project provinces are similar to those in non-project provinces
and suggest that the project was not effective in improving the use of contraceptives.
This may be partly due to the lower emphasis on family planning support when the
project was refocused on MNCH-related activities (paras. 3334).
2.

Guaranteed access of the poor to essential


essential health and family planning
services

78.
ADB review missions noted that the project did not help find strategies to
strengthen propoor targeting (of ongoing social safety-net schemes such as the health
card), and out-of-pocket expenditure on health care for this population segment
remained high. It was also highlighted that the poor often need to travel long
distances, and the project did not help find mechanisms that could cover opportunity
and transport costs. The evaluation examined public health facility visit for the poorest
quintile using National Socioeconomic Survey (SUSENAS) data and noted large
increases between 2001 and 2007 were reported for Riau, where rates increased from
3.0 to 11.9 visits per 100 people. Other provinces with large increases were Riau
Islands, North Sulawesi and Southeast Sulawesi, where the rates almost doubled.
As already noted, a large number of factors contributed to the observed increase in
facility use. Between 1997 and 2007, overall use of both inpatient and outpatient
health services among those who report illnesses increased (footnote 9). Jamkesmas,
the government-funded health insurance scheme for the poor, covered about 40% of
poor and near-poor households.27 The project supported implementation of the reform
strategies such as Jamkesmas and financed investments to improve lower-level public
health infrastructure, and provided basic equipment most often used by medical staff
for diagnosis of the poor.
79.
Inspection of these facilities by the evaluation team indicated that most had a
functioning basic set of diagnostic equipment and basic pharmaceutical supplies. When
asked how the poor had benefited from the project, most stakeholders told the
evaluation team that insurance and improvement in infrastructure had been key
factors. Some also highlighted that the project had directly targeted investment in
lower-level public health centers most frequently used by the poor.
80.
Given that outpatient contact rates in project and non-project provinces do not
appear to differ substantially, the evaluation did not explore differential use by wealth
quintile. Exploration of the incremental impact of the project was investigated primarily
for the economic reevaluation, in which per-quintile data is not considered. Given the
increased facility use by the poorest quintile, the evaluation concludes that the project
met its health facility targets on access and use by the poor. In the absence of quintile

27

P. Harimurti et al. 2013. The Nuts & Bolts of Jamkesmas Indonesias Government-Financed
Health Coverage Program. World Bank. Washington, DC.

21
Several family
planning
indicators are
stalling and
even falling in
some provinces

22

Decentralized Health Services Project in Indonesia


analysis per project and non-project district, it is difficult to quantify the magnitude of
the impact that could be attributed to the project.
3.

The overall
upward trend
trend
in the
allocation of
local budget
to health is
evident

District budget for health

81.
A target in the report and recommendation of the President was that most
district governments would allocate at least 4% of their budget to health. It is evident
that all provinces achieved this target in their own budgets, but there was no
systematic collection of data on how far the districts themselves achieved this.
The same issue was highlighted in completion reports for the World Bank's provincial
health projects. Like many indicators assessed in this report, the estimates have a great
deal of year-on-year variation (Table 5). For example, the percentage allocation of Balis
budget to health increased from 3.5% in 2002 to 9.3% in 2005, then dropped to 3.9%
in 2011. The definition of the indicator appears to differ on whether to include or
exclude the value of health insurance. Such variation makes specific time series
evaluation impractical. Despite this issue, the overall upward trend in the allocation of
local budget to health is evident, particularly in the provinces with a larger
concentration of the poor (NAD, Riau, Riau Islands, Bengkulu, Central Sulawesi, and
Southeast Sulawesi.
Table 5: Average District Budget by Province Allocated to Health
(Percentage of total expenditure)
Province
NAD
Riau
Riau Islands
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi

2002
(%)
3.5
3.2
n.a.
4.6
3.5
7.9
0.7
1.7

2003
(%)
4.2
3.2
n.a.
1.4
3.8
16.9
3.7
3.0

2005
(%)
n.a.
3.8
3.0
5.6
9.3
3.8
6.9
6.9

2007
(%)
8.8
4.I
4.8
n.a.
7.2
5.7
7.9
8.6

2011
(%)
10.3
6.2
7.4
13.5
3.9
5.7
11.6
19.7

NAD = Nanggroe Aceh Darussalam, n.a. = not available.


Source: Data collected by the evaluation team.

District level
appropriate
planning and
budgeting
tools are used
in preparing
district health
plans

82.
A review of district health expenditure in 2006 found that at least 40% of
public expenditure on health is for personnel, and districts have limited discretion over
the deployment of public expenditure on health.28 Despite decision making being
limited, the evaluation team observed that at the district level appropriate planning
and budgeting tools are being used in preparing district health plans. Low levels of
public spending on health have been deemed to be a key reason for poor health
indicators, such as high mortality rates and the excessive use of self-treatment in
Indonesia. Key informants indicated that training provided under the project has
assisted in mobilizing local level resources for the health sector.
83.
The project has also helped promote counseling services and awareness of
health issues among both men and women. However, the outreach of family planning
services has stagnated in recent years due to the reduced focus in the later stage of
project implementation. The PCR noted that perceived risks associated with side effects
28

P. Heywood and N. Harahap. 2009. Public funding of health at the district level in Indonesia
after decentralization sources, flows and contradictions. Health Research Policy and Systems
2009. No. 7:5.

Performance Assessment

23

of contraceptive use might have contributed to unmet needs29 for family planning
services. The decrease in family planning indicators might be linked to the decrease in
BKKBN budget and the reorganization of BKKBN.
84.
The PCR also acknowledged that the government acted to mitigate the impact
of the Asian financial crisis in 1997 by providing block grants to puskesmas and
midwives to support local health services. Adoption of Askeskin (a health insurance for
the poor) from 2006 has enabled provision of free outpatient and limited inpatient care
for the poor, and nutritional supplements for pregnant women and newborns.
The coverage of the insurance scheme varies across the provinces. Nevertheless, there is
no evidence of significant differences in health service use between project and nonproject provinces.

D.

Efficiency

The government
acted to
mitigate the
impact of the
Asian financial
crisis in 1997 by
providing block
block
grants

85.
Like the PCR, the evaluation rates the project's performance less efficient.
There were significant delays at startup because implementation units had limited
experience with ADB operational procedures. ADB received the governments decree for
establishing the central project management unit, 30/MENKES-KESOS/SK/2001,
on 11 January 2001, but it became effective only on 25 June 2001. This was largely
associated with delays in the issuance of legal opinion on project implementation
before all necessary legal documents between ADB and the Government of Indonesia
had been endorsed.
86.
The problem was exacerbated by the need to establish a large number of
implementation units when Indonesia adopted decentralization and devolution of
decision making to the districts. Since all districts in each project province were eligible
for support, more than 50 districts had to prepare investment plans during the first
phase to program civil works, equipment, and much of the training budget. No TA was
provided at startup.
87.
Given the perceived complexity of ADB procedures and limited project
preparation, the omission of TA compounded the lack of international project
management experience. Although delays were inevitable given the large changes in
organizational structures and processes within the public health service during
decentralization, these could have been reduced to some degree with appropriate TA.
Project phasing should have reflected this degree of complexity, devoting more time to
capacity development in the initial 2 years. Also, as already mentioned, the number of
project provinces and districts could have been reduced to manage the demand for
health service capacity development.
88.
The PCR noted that 55% of the loan period had elapsed by the end of 2003,
when the midterm review was conducted, and only 21% of contract awards and 16%
of disbursements had been undertaken. During the later stages of the project,
implementation was slowed by late provision of counterpart funds, which repeatedly
narrowed the implementation window.
89.
Other factors that undermined project efficiency included (i) transfer of large
numbers of health service staff to districts, but not all districts were willing to take
them on board and instead appointed recruits from within their districts; (ii) low
29

Unmet need is the statistical measure that calculates how many sexually active women say they
want to stop or delay child bearing.

The omission of
TA
compounded
the lack of
international
project
management
experience

24

Decentralized Health Services Project in Indonesia


absorptive capacity of district health offices; and (iii) long delays in release of
counterpart funding from the Ministry of Finance.30 As a result the project had to be
extended twice.
90.
Implementation arrangements. At the completion of the project, eight
provincial implementation units and 73 district implementation units had been
established. A technical review team was set up in every province to review district and
provincial proposals and annual reports. The team comprised technical experts from
provincial health services and BKKBN, the private sector, the community, professional
health organizations, local academic institutions, and project consultants.
ADB supervision missions noted that BKKBN staff periodically participated in the
provincial and district meetings but since they were not involved in local situation
analysis and in the selection of family planning activities, their input at the meetings
was limited. The PCR noted that a central technical committee (jointly staffed by MOH
and BKKBN) for the family planning component of the project did not exist, and this
may have constrained coordination during project implementation.

The
The creation
of new
districts in the
project area
reduced
allocations of
resources for
local activities

91.
The project was also designed with a view that health service delivery would be
devolved to the provincial authorities. Resources for district-based project management
and associated capacity development were therefore limited. This oversight largely
reflected the evolving nature of decentralization and the ensuing uncertainties at the
time of project formulation and inception (footnote 11). The creation of new districts in
the project area reduced allocations of resources for local activities, which contributed
to delays in implementation. It was agreed that project resources originally allocated
for physical investments (health facilities and equipment) in the new districts would be
maintained but managed by the district implementation unit of the "old" district or by
the provincial project implementation unit. This arrangement complicated districts
implementation of project activities. Limiting the number of districts, or allocating
more resources to those already included in the design, would have avoided this issue,
and would have led to more efficient implementation.
92.
Economic
Economic reevaluation. The project economic analysis indicated an overall
benefitcost ratio of more than 4:1. The PCR, however, did not present economic or
financial analysis. This evaluation has reevaluated the appraisal estimate using revised
burden-of-disease estimates and the economic value of labor productivity losses
avoided as a result of improved medical services.
93.
The evaluation teams interviews with key respondents resulted in updated
(2010) burden-of-disease estimates for Indonesia and a longer cost-benefit time frame.
Key benefits stem from decreasing mortality rates in provincial surveys in 2007 and
2012 undertaken by IDHS and captured in the most recent burden-of-disease
estimates.31 When non-project provinces are considered, the difference in mortality
rates and coverage of key health services (outpatient contact rates in health centers,
30

31

In 2006, budget release was delayed by 8 months and hence implementation of project
activities had to be squeezed into 4 months.
Statistics Indonesia (Badan Pusat StatistikBPS) and Macro International. 2008. Indonesia
Demographic and Health Survey 2007. Calverton, Maryland: BPS and Macro International;
Statistics Indonesia (Badan Pusat StatistikBPS), National Population and Family Planning
Board (BKKBN), and Kementerian Kesehatan (KemenkesMinistry of Health), and ICF
International. 2013. Indonesia Demographic and Health Survey 2012. Jakarta: BPS, BKKBN,
Kemenkes, and ICF International; and Institute for Health Metrics and Evaluation, Human
Development Network, The World Bank. 2013. The Global Burden of Disease: Generating
Evidence, Guiding Policy East Asia and Pacific Regional Edition. Washington, DC.

Performance Assessment

25

vaccination coverage, use of health care for childhood diseases) are not significant.
Correspondingly only a minor contribution of the project to improved health outcomes
can be estimated at the provincial level. Appendix 3 contains detailed analysis using a
time frame of 20 years, which more closely reflects the economic life of the
infrastructure developed under the project.
94.
It is estimated that the project requires a reduction in burden of disease in
project provinces of 2%. The breakeven burden-of-disease reduction is less than 4%
estimated at appraisal. Although comparison of health indicators in project and nonproject provinces found no evidence of burden of disease decrease, the evaluation
team, in discussions with key informants, noted that project has had positive health
benefits. Given only marginal decrease in burden of disease is required for breakeven, it
is likely that the project generated marginal net benefits.

Discussions with
key informants,
noted that
project
project has had
positive health
benefits

95.
Monitoring and evaluation. MOH supported the compilation of benefit
monitoring reports at the end of the project for all project provinces. In each province,
two districts (or cities) with high and low Indonesian Human Development Index 2002
ratings were selected to measure outcomes in terms of human development: longevity
(life expectancy), knowledge (education attainment), and a decent standard of living
(adjusted income). Exit interviews were also used to assess clients satisfaction in a
sample of 30 respondents per health center. The scores of the survey were provided in
the English version of the report. However, the instrument and samples used to
generate scores were not provided. Gender-disaggregated data was collated in regard
to program inputs and a range of case studies conducted to illustrate selected impacts.
96.
Reporting during implementation was not satisfactory. Within 6 months of
loan effectiveness the central project management unit was to develop a performance
monitoring system for training and capacity-building activities, and each province was
to develop an M&E framework for local activities. TA was to be provided to support
collection of data. At midterm, ADB requested that M&E must be strengthened. Many
districts had not yet identified indicators related to their local priorities, training quality
was not systematically assessed, and the health information system was not developed.
As a result quarterly implementation reports were often delayed and not
comprehensive. To systematically assess impacts, facility-level data from project and
non-project provinces needs to be analyzed. Additionally, a survey of health outcomes
using a statistical approach is required to quantify impacts. Indicators such as maternal
and infant mortality are not likely to be meaningful at this level, given the low numbers
of cases relative to population size. Indicators for maternal health such as obstetric
complications, referrals, client satisfaction, and health expenditures are more likely to
capture any differences in the health status, and thus project impact.

E.

Sustainability

97.
The project is likely to be sustainable. The evaluation team found during its site
visits that buildings and equipment were in good condition and well used.32 Almost all
provincial and district health officials interviewed considered ongoing financial support
from local government resources to be sufficient for the current capacity and coverage
of public health facilities.

32

Reporting
during
implementation
was not
satisfactory

With the exception of the provincial hospital of Central Sulawesi, which has been relocated to
another area that is larger in size and not prone to flooding.

Local
governments'
budgets for
health had
increased

26

Decentralized Health Services Project in Indonesia


98.
The evaluation team and the PCR both noted that local governments' budgets
for health had increased as a percentage of overall district budgets in most of the
districts with high incidence of poverty. Local government regulations and decrees have
also been issued to ensure the sustainability of health and family planning investments
and the continuation of health-care reforms.
99.
The evaluation team witnessed an ongoing district health planning and
budgeting exercise at visited health departments. Various other components of the
project have been institutionalized and were being supported by national and local
organizations. Subsidy schemes for traditional birth attendants continue and have been
replicated in other areas. Awareness of maternal care has improved, along with basic
emergency obstetric care, and emergency obstetric care referral systems have been
accepted and continue to expand beyond project completion.

Referrals have
more than
tripled in
several project
provinces

100.
The degree of increase in obstetric and neonatal emergency referrals is evident
in Table 6. Referrals have more than tripled in several project provinces. NAD in
particular has seen the largest percentage increase in obstetric referrals, from less than
2,000 cases in 2003 to more than 22,000 in 2010. While other investments have flowed
into health services beyond this project, it is evident from the stakeholders that ADB
funding has played a role in strengthening health services in several districts.
Institutionalization of health insurance for the poor has been recognized as a major
achievement under the project and good practices identified during project
implementation have been adopted in the follow-on DHS2 (footnote 13).
Table 6: Obstetric and Neonatal
Neonatal Emergency Referrals by Province

Province
NAD
Riau
Bengkulu
Riau Islands
North Sulawesi
Central Sulawesi
Southeast Sulawesi
Bali

Obstetric Emergency Referrals


2006
2010
2003
1265
1147
22,263
6536
3791
27,466
4857
678
7,942
NA
907
9,994
3057
1467
9,706
n.a.
1754
12,472
2228
2068
9,552
3336
1418
14,085

Neonatal Emergency Referrals


2003
2006
2010
49
850
14,832
6,886
2,621
19,119
716
374
5,526
n.a.
267
6,374
426
163
6,646
6,230
442
7,387
2,035
244
5,984
1,036
615
9,697

n.a. = not available.


Source: Independent Evaluation Departmentdata collected from the Ministry of Health and provincial
health offices.

Good
practices
identified have
been adopted
Staff trained
under the
project are
now in
influential
positions

101.
The major thrust of the project-supported policy reforms, including health
insurance for the poor and subsidies for traditional birth attendants, is likely to be
sustained. ADB assistance in the sector and government commitment helped prioritize
universal coverage in provinces such as Bali. The project provided some continuity to
previously established programs, in particular those related to the block-grant funding
of health centers under the governments earlier social safety-net program. MOH has
made an effort to ensure continuity of the project outputs by incorporating these
elements into other aid-agency-supported activities on maternal health. Good practices
identified during the project implementation have been adopted in DHS2
(footnote 13).
102.
Furthermore, technical and managerial staff trained under the project are now
in influential positions in the governments health system and are better placed to
ensure that good practices under the project are implemented effectively and sustained
for the benefit of people, particularly those in geographically challenging areas.

Performance Assessment

103.
At the district level, most of the health service facilities still rely heavily on
annual budgetary allocation from the central government, while local government
budgets are used for hiring staff for the health facilities, and purchasing equipment
and medical supplies. At times, allocation from the central government, particularly for
health infrastructure improvement or expansion, tends to be inadequate to service
demand for health services.

27

CHAPTER 4

Other Assessments
104.
Chapter 3 was devoted to the assessment of four core evaluation criteria:
relevance, effectiveness, efficiency, and sustainability. This chapter discusses on other
evaluation parametersproject impacts, performance of ADB and the borrower, role of
development partners in Indonesias health sector, and the performance of technical
assistance to the project.

A.

Impacts

105.
This section is based on data analysis, largely sourced from IDHS published in
1997, 2003, and 2007 and findings from the 2012 IDHS. Although the PCR was
prepared in 2010, it largely relied on the 2007 IDHS. The 2012 report provides a more
recent picture of the health status in Indonesia and helps understand health outcomes
and health facility utilization, in both project and non-project provinces.
1.

The project
has helped
Indonesia to
stay on track
to achieve
reduction of
child mortality

Selected Millennium Development Goals

106.
Infant mortality rate. The project has helped Indonesia to stay on track to
achieve one of the key health-related MDGsreduction of child mortality. Performance
targets included reducing the U5MR and infant mortality rate (IMR) to at least 30%
below the local benchmark data. Reducing the U5MR by two-thirds between 1990 and
2015 is the MDG 4 goal. Provincial U5MR data provided by the government during the
evaluation showed decreases in all project provinces except NAD, and Southeast
Sulawesi (Table A4.1). Variations in year-on-year provincial data were evident for this
indicator.
107.
Nationally, the IMR has shown a large, although slowing decline since 2003.
This trend follows what had been evident since the 1990s, as demonstrated in Figure 3.
However, Indonesia's IMR is still higher than in other countries within the Association of
Southeast Asian Nations (e.g., the Philippines, Thailand, and Viet Nam).

Other Assessments

29

Figure 3: Infant Mortality


Mortality Rate
(per 1,000 live births)
120
100
OECD
80

Vietnam
Indonesia

60

Philippines
Cambodia

40

Laos
20

Thailand

0
1995

2000

2005

2010

Source: Independent Evaluation Mission data.

108.
In Indonesia, the overall IMR declined in the target provinces from around
42.1 deaths per 1,000 live births in 2003 to 37.0 in 2012 (Appendix 4, Table A4.1).
However, significant variations are noted across the provinces. For example, in Riau and
Bengkulu provinces the downward trend has been consistent between 2003 and 2012,
while other provinces have experienced increases between the survey years. The IMR
increased in Bali and North Sulawesi between 2003 and 2007, but then decreased in
2012.33
109.
While the overall IMR decreased in Indonesia between 2003 and 2012,
very small differences in change have been observed between the project and nonproject provinces. For example, in 2003 the IMR averaged 42.1 in project provinces and
43.2 in non-project provinces, which declined to 37.0 and 36.3 in 2012, respectively.
The overall difference between project and non-project provinces was greater in 2007,
largely because of changes in provinces included in 2003 and 2007 in IDHS, but the
difference narrowed significantly in 2012. These findings also demonstrate the
instability of provincial mortality results. The use of longer-term moving averages
hinders the comparison of provincial indicators because the results of survey years prior
to project implementation are included.
110.
The IDHS reported that children in richer households had lower mortality rates
than those in poorer households in 2007. For example, the IMR for children in the
lowest wealth quintile was 56 deaths per 1,000 live births, compared with 26 deaths
per 1,000 live births for children in the highest wealth quintile. The observation
highlights the need to improve access to and use of health services by poor people.
Declining immunization coverage and still widespread pockets of malnutrition may be
hindering further progress in the improvement of this indicator.
33

It should be noted that because fertility levels are low in Indonesia, the IDHS infant and child
mortality estimates are based on relatively small numbers of cases, which leads to unstable
estimates. To reduce this problem, mortality measures based on the 2007 IDHS are calculated
for 5-year birth interval periods.

Significant
variations are
noted across
the provinces

30

Decentralized Health Services Project in Indonesia

111.
Maternal mortality ratio. Survey respondents cited training, health reforms, and
infrastructure improvements carried out under the project as vital in improving
maternal health outcomes. The national MMR34 in 2000 was estimated to be 353 with
a wide confidence interval, which decreased to 220 in 2010. However, attributing gains
to the project is made difficult because the rate is based on a 5-year moving birth
cohort. Low fertility in Indonesia skews statistics, given the relatively small number of
births per year relative to the total population. The government set a target for the
number of pregnant women accessing reproductive health services to be 90% by 2010.
This target was included in the project objectives, but has not been met in most
provinces, with the exception of Bali.
2.

Maternal Health
Health and Gender Impact

112.
Overall, the project has had an impact on womens health. It developed a
subsidy scheme for traditional birth attendants to give pregnant women skilled
antenatal care and delivery, which continues and has improved the overall coverage of
skilled health care for women. As such, the project played an important role in raising
awareness about womens reproductive health issues and incorporating gender in
community health services.

Midwives cited
a lack of
capacity in
obstetrics
obstetrics and
the delay in
referral
treatments as
main reasons
for maternal
death

113.
Access to caesarean section for childbirth. During the evaluation teams
discussions, midwives cited a lack of capacity in obstetrics and the delay in referral
treatments as main reasons for maternal death. Health professionals received technical
training in reproductive health, and local government officials raised awareness of the
importance of maternal health in local communities. As a result, the proportion of
women opting for caesarean section is increasing in all project provinces (Appendix 4,
Table A4.2). Large increases are evident since 2006 in more prosperous provinces such
as Riau and Bali. In Bali, women opting for caesarean section increased from 1.7 % of
all deliveries in 2006 to 14.1% in 2010.
114.
Assistance during child delivery. The government had set a target of 90% of
childbirths to be assisted by medical staff by 2010. The project also adopted this target.
The coverage of skilled health staff increased from 69.1% in 2003 to 83.7% in 2012 in
project provinces (Appendix 4, Table A4.3). This was slightly higher than in non-project
provinces (from a population-weighted average of 68.1% in 2003 to 82.1% in 2012).
Only Bali and NAD achieved the national target. Central Sulawesi and Southeast
Sulawesi are far behind other provinces (63% and 66%) but it should be recognized
that these provinces started from a low base of 54% and 42%.
115.
Antenatal care. In Indonesia, antenatal care is defined as pregnancy-related
health care provided by a medical professional.35 Its coverage has steadily increased.36
Central Sulawesi, NAD, and Southeast Sulawesi had major increases in access to
antenatal care (Appendix 4, Table A4.4). Overall, the population-weighted
achievements have been similar in both project and non-project provinces. Since only
women who gave birth in the last 5 years responded to the IDHS question on the use
of antenatal care, full project impacts are yet to emerge.

34
35

36

The maternal mortality rate refers to the number of maternal deaths per 100,000 live births.
Medical professionals include general practitioners, obstetricians, gynecologists, nurses, or
midwives.
Women who had a live birth in the 5 years before each IDHS were asked who provided
antenatal care during their pregnancy.

Other Assessments
3.

31

Child Health

116.
Child immunization. Vaccines are provided to health service centers through
the national and provincial health systems. Children receive vaccinations from village
health posts (posyandus), maternity clinics (polindes), and health centers (puskesmas).
Project support for the development of these facilities may have contributed to better
immunization coverage.37 Provincial data suggests that immunization coverage
dropped in 2007 in at least five of the project-supported provinces compared with
2003 levels, while North and Southeast Sulawesi experienced some increase (Appendix
4, Table A4.5). According to the 2012 IDHS, compared with 2007, all project provinces
have realized greater coverage, i.e., above the 2007 overall level (up from 52.4% to
65.0%). The immunization coverage in non-project provinces increased from 61.7% to
69.7%. This may have been driven by Indonesia launching a vigorous national
immunization drive after 2007.
117.
Diarrhea treatment at health service facilities. Diarrhea can be fatal, particularly
for children, if it is not treated in time. The IDHS asked mothers of children who had
diarrhea whether the children had been taken to a health service facility.38
Data confirms that the use of health facilities for diarrhea treatment increased from
57.1% in 2007 to 65.9% in 2012 in project provinces (Appendix 4, Table A4.6),
although to various degreesBengkulu led with 81.6% in 2012 while Southeast
Sulawesi lagged at 55.4%. Between 2007 and 2012, Bali was the only province with
lower use of service facilities.

Use of health
facilities for
diarrhea
treatment
increased from
57.1% to 65.9%

118.
Nationally, average health facility utilization increased from 51.0% in 2007 to
64.6% in 2012, against an increase from 57.1% in 2007 to 65.9% in 2012 in project
provinces. The empirical evidence from IDHS is consistent with Indonesias SUSENAS
results.
4.

Family Planning

119.
The project also aimed to improve family planning services in project areas.
Unintended pregnancies, at an estimated 24 million per year, are a major issue in the
country. Many of these women seek abortion, which in turn increases the rate of
maternal mortality. Jakarta Post quotes BKKBN data and reports that the number of
abortions in Indonesia increased by 15% annually; and in 2012, 2.4 million abortions
took place in the country, of which about 800,000 were teenagers.39 Several sources
report that many abortions are carried out under unsafe conditions and at great risk.
120.
Key informants involved in delivering family planning training under the project
told the evaluation team that the training largely focused on disseminating and sharing
knowledge about birth control implants and use of injections. Although the family
planning component faced implementation difficulties due to the impact of
decentralization, better family planning and lower incidence of unwanted pregnancies
have had a positive gender impact in that more women are aware of available options.

37

38

39

Children are considered fully immunized when they have received one dose of vaccine against
tuberculosis (BCG), three doses each of the DPT (diphtheria, pertussis, and tetanus) and polio
vaccines, and one dose of measles vaccine.
Refers to children under the age of 5 who had diarrhea in the 2 weeks preceding the survey
and were taken to a health facility.
Available:
http://www.thejakartapost.com/news/2013/02/20/abortion-today-still-secret-easyfind.html, as of 25 December 2013.

Many abortions
are carried out
under unsafe
conditions and
at great risk

32

Decentralized Health Services Project in Indonesia


121.
Contraceptive prevalence rate among married women (1549 years).
IDHS reports data on the prevalence of contraceptive use among never-married women
and currently married women aged 1549. This indicator is used for measuring the
success of family planning programs. The results indicate that overall there has not
been a major shift in the prevalence of contraceptive use between 2003 and 2012
(Appendix 4, Table A4.7). Some improvements were seen between 2003 and 2007 in
selected provinces, but these gains reversed in 2012. The prevalence rates were mostly
similar in project and non-project provinces but collectively non-project provinces had a
marginally higher rate than the project provinces (62.4% vs. 59.0%). Among the project
provinces, NAD had the lowest prevalence rate in both 2007 and 2012.
122.
Unmet need for family planning services. Unmet need for family planning has
modestly increased in both project and non-project provinces (Appendix 4, Table A4.8).
Available data show that it declined marginally between 2003 and 2007, largely due to
Riau, Bengkulu, Bali, Central Sulawesi, and Southeast Sulawesi. However, it increased in
all provinces in 2012, compared with 2007.

B.

Performance of ADB, Borrower and Development Partners


Partners
1.

The evolving
decentralization
decentralization
required close
supervision of
the project, and
adjustments
both from ADB
and the
government

ADB Performance

123.
Overall, ADBs performance is rated satisfactory. As listed in the PCR,
ADB fielded 13 project review missions, 1 emergency rehabilitation mission, 1 midterm
review mission, 1 final review mission, and 1 PCR mission. The evaluation considers that
a two-part midterm review mission between December 2003 and March 2004
responded to the needs of the project and country priorities. While implementation
progress throughout the project period was considered satisfactory or highly
satisfactory, the evolving decentralization required close supervision of the project, and
adjustments both from ADB and the government. ADB also positively responded to
evolving organizational change within MOH and placement of the project
implementation responsibility as desired by the government. ADB also approved two
extensions of the project (24 months and 3 months) at the request of the government,
which helped complete project activities as much as possible.
124.
The evaluation team, through focus group discussions and key informant
interviews, learned that ADB review missions largely concentrated on Jakarta and took
only limited field visits. According to the PCR, ADB also changed project officers four
times and project analysts twice. ADB staff turnover slowed communication between
project and ADB staff. Nevertheless, the government and project staff deemed ADBs
support during implementation satisfactory.
2.

Borrower Performance

125.
The evaluation rates the borrowers performance less than satisfactory under
the decentralization model adopted by Indonesia. The PCR noted that the project was
managed by different executive secretaries in different departments, with limited
integration between health and family planning activities. The midterm review
highlighted that a major issue in project administration was the lack of timely
systematic reporting of project activities, making it difficult to react swiftly. The main
reason for this significant delay was the need for consolidation of district reports at
central and provincial levels, along with many districts having a poor understanding of
project administration.

Other Assessments

33

126.
The government did not provide timely counterpart funds during the project
period, although local governments allocated sufficient funds for sustaining civil works
and equipment. This led to a much shorter implementation period in a fiscal year,
which may have compromised the quality of work completed. The Financial and
Development Supervisory Agency (Badan Pengawasan Keuangan dan Pembangunan)
issued an audit report in 2007 that the project had ineffective supervision and review
of activities. The audit report enumerated nine cases amounting to Rp691,762,960 and
involving (i) overpayment on training and workshop per diems, civil works contractors,
suppliers, and consultants (MOH/BKKBN); (ii) poor supervision of civil works (MOH);
non-use of medical equipment and completed civil works for the intended purpose
(MOH and BKKBN); and ineligible payments to fellowship participants (MOH).

The
government
did not
provide timely
counterpart
funds

127.
The evaluation findings confirm the PCR statement that formal coordination
was lacking between MOH and BKKBN, an agency responsible for family planning
services. As a result, during the midterm review the focus was mostly on the provision
of MNCH services and less on family planning. Even for health service delivery, project
implementation was largely seen as being driven by MOH, with limited ownership by
the local and district health offices.

Formal
coordination
was lacking
between MOH
and BKKBN

128.
Discussions with different stakeholders also indicated that initial
implementation delays were largely due to inadequate local and district institutional
capacity, which was to be supported by engaging services of a project management
firm. The evaluation team noted that this did not happen and that the project opted to
engage a small number of individual consultants. Furthermore, as stated in the PCR,
the project implementation units were managed by retired civil servants with no
experience in health services decentralization and ADBs operational policies and
guidelines. Also, it was reported that project staff turnover was quite high.
129.
The borrower should have emphasized that the project needed a baseline
survey so that project progress could be monitored over time. However, this did not
happen until after the midterm review, which prevents systematic analysis of project
impact. The TA team finally helped set up a health indicator monitoring system by
incorporating selected MNCH indicators.
3.

Development Partners Performance


Performance

130.
No other development partners were involved in the project, although other
agencies also supported health reform and associated activities in Indonesia. The World
Bank, Japan Bank for International Cooperation, United States Agency for International
Development, Australian Agency for International Development, the Netherlands, and
specialized United Nations agenciesUnited Nations Development Programme, United
Nations Childrens Fund, United Nations Population Fund, and World Health
Organizationhave been active in the support of Indonesia's health sector. External
assistance has averaged $200 million$350 million per year between 2000 and 2010,
or around 1%2% of total health expenditure.40
131.
The World Bank financed a series of provincial health projects with similar
objectives to those of the DHS. One was the Provincial Health Project in Lampung and
Yogyakarta, which was implemented between 2000 and 2007 with World Bank

40

World Health Organization. 2013. Global Health Observatory


Available: http://apps.who.int/gho/data/ node.country.country-IDN

Data

Repository.

34

Decentralized Health Services Project in Indonesia


financing of $31.26 million.41 It included activities to support newly decentralized
organizational structures, health financing mechanisms, workforce policies, and local
health service delivery and financing arrangements. The independent completion report
(footnote 12) rated the project moderately satisfactory, but noted that the central
government had limited capacity for decentralization, project performance indicators
required better definition, and provinces and districts needed more guidance.
The report emphasized that the enabling environment provided by the central
government had significant shortcomings for project implementation from the point of
view of lateness in disbursement of funds and confusion over precise allocation of
authorities among government levels.
132.
The Second Provincial Health Project was implemented between 2002 and 2007
in the provinces of North Sumatra, West Java, and Banten. It had $57 million in World
Bank financing and the objectives of assisting effective health sector decentralization,
initiating sector reforms, and protecting health services for the poor. Components were
designed to tackle decentralization issues in districts and provinces, and to assist
communicable disease control, equity in service provision, establishment of a food and
drug control agency, and supporting health research. The project was rated as having
an unsatisfactory outcome. Implementation was constrained by districts not having
clearly defined functions and standards, poor performance indicators, along with the
scope of the project being too broad and attempting to implement too wide a scope
of reforms at the same time placing a large burden on project implementers.42

Complex
projects
require
considerable
capacity
capacity
development

133.
From 2003 to 2008, the World Bank implemented the Health Workforce and
Services Project.43 It aimed to support health sector decentralization in Jambi,
East Kalimantan, West Kalimantan, and West Sumatera through improved health
financing and services, and to strengthen health workforce policy and management.
The completion report rated the project outcomes moderately unsatisfactory. Lessons
included that complex projects require considerable capacity development during the
early years of implementation and warrant a longer implementation period, and that
M&E responsibilities should be made clearer at startup. The most difficult obstacle to
implementation, however, was lateness in the central governments disbursement of
project funds.
134.
The United States Agency for International Development has focused on
infectious diseases, including tuberculosis, HIV/AIDS, neglected tropical diseases, avian
and pandemic influenza, along with maternal and child health.44 In the past the agency
was a major supporter of family planning in Indonesia, but the country has since
graduated from being eligible for assistance. In the field of maternal health, the
program includes activities to support obstetrical care and referral systems, and care for
children suffering respiratory illnesses. Indonesia has also been a large recipient of
Global Fund grants to support prevention of HIV/AIDS, malaria, and tuberculosis, and
associated care. Since 2003 some 17 grants valued at $500 million have been targeting

41

World Bank. 2000. Provincial Health Project in Indonesia. Washington, DC.


Available: http://documents. worldbank.org/curated/en/2000/05/437378/indonesia-provincialhealth-project
42
World Bank. 2008. Second Provincial Health Project in Indonesia. Washington, DC.
Available: http://documents.
worldbank.org/curated/en/2008/06/9735922/indonesia-secondprovincial-health-project
43
World Bank. 2003. Health Workforce and Services Project in Indonesia. Washington, DC.
Available: http://documents.
worldbank.org/curated/en/2003/05/2216140/indonesia-healthworkforce-services-project
44
GHI. 2011. Indonesia GHI Country Strategy.

Other Assessments

35

these diseases. The grants contain funds to support health worker training, community
outreach, capacity of laboratories, vector control, testing, procurement of insecticidetreated nets, and enhancing the health information system.
135.
The Global Fund and the Australian Agency for International Development are Capacity
currently the major supporters of health systems strengthening. The Global Fund health developed
systems strengthening program entails a grant of $16.5 million to develop a national
under the DHS
health information system, scaling up the sample registration system, and promoting
health system research. The Australia Indonesia Partnership for Health Systems is now being
Strengthening (valued at $50 million from July 2011 to June 2016) will link with the deployed in
Global Fund program. It aims to expand the number of qualified nurses and midwives,
other donordonorand support local governments to staff health centers. Additionally, the Australia
Indonesia Partnership for Maternal and Neonatal Health (valued at $16.6 million in supported areas
June 2012 aims to increase the number of births covered by skilled birth attendants,
increase the number of women using health facilities to give birth, and help reduce
malaria and anemia among mothers. Capacity developed under the DHS is now being
deployed in other donor-supported areas, and lessons and experiences from these
projects should be disseminated. This issue is discussed in chapter 5.

C.

Technical Assistance

136.
Attached to the loan was TA to assist MOH in defining, implementing,
and evaluating health sector reforms in the context of decentralization. The TA had
six outputs: (i) health sector reforms identified, (ii) capacity for health system
management and health service delivery built, (iii) health plans developed,
(iv) awareness of health priorities created among decision makers, (v) operations
research capacity developed, and (vi) project management supported.
137.
Originally, $1 million in TA was approved, but this was revised during
implementation to $1.86 million, with the Australian Agency for International
Development cofinancing an additional $0.86 million for consulting services and
operational costs. The TA supported management and planning training, and
assistance with health sector reforms. The scope was broadened to include tsunami
damage assessment and reconstruction, national health accounts, and strengthening of
MNCH. The TA was extended four times and consultants replaced due to poor
performance.
138.
The TA had the same overall objective of improving the health status of the
population in the project areas, and of assisting the planning of health reforms under
objective 2. According to MOH staff interviewed during the evaluation, the TA
supported the governments intention to strengthen local capacity for health service
delivery. Additionally staff noted that the TA helped introduce initiatives like health
insurance, health financing, drug use management, health-care quality improvement,
and publicprivate partnerships. Training modules developed under the TA helped train
more than 200 district health staff on evidence-based health planning and budgeting
in project and non-project districts. In two instances, the evaluation team witnessed a
planning and budgeting exercise, and staff's active participation was evident. The TA
also helped MOH assess damage and reconstruction efforts after the 2004 tsunami in
NAD. The TA completion report rated the TA satisfactory.
139.
The PCR states that the TA helped the project produce several documents on
different health sector initiatives and training modules in health planning and
budgeting; however, health sector reform [was] not fully achieved because the focus

Training
modules
developed
under the TA
helped train
district health
staff on
evidenceevidence-based
health planning
and budgeting

36

Decentralized Health Services Project in Indonesia


was largely on documentation of existing reforms rather than launching and evaluating
new local initiatives. The evaluation team could access only a few of the outputs
produced under the TA and, hence the quality of consultants outputs could not be
assessed. Moreover, several outputs had no English translation. The consultant
recruitment approach provided flexibility in engaging individual consultants, but
coordination of various outputs proved challenging for project management.
The evaluation considers that sequencing of consultants engagement could have been
more strategic. For example, district health planning could have been followed by M&E.
Moreover, tangible contribution from the TA came only after the midterm review. If a
firm had been engaged from the beginning, some project implementation delay could
have been avoided. In view of the broad geographic scope of the project and the
technical nature of decentralization-related challenges, the TA was originally underresourced.

CHAPTER 5

Issues, Lessons, and FollowFollowUp Actions


140.
This chapter draws on evaluation findings discussed in preceding chapters.
It identifies key issues facing Indonesias health sector, lessons derived from the project
design, implementation and completion; and recommends follow-up actions by ADB
and the government.

A.

Issues
1.

Geographic Targeting

141.
According to senior staff at MOH, the original selection of seven provinces
across Indonesia intended to avoid duplication of resources between ADB and the
World Bank. The rationale for this selection, as provided to the evaluation team,
was weak. These provinces did not have proportionately larger poor populations,
particularly adverse health indicators, or several districts without doctors, nor were they
experiencing issues with overall economic development. Instead, the selection
generated a very broad and scattered footprint for implementation. Some of the
provinces, such as Bali and North Sulawesi, have very favorable health indicators.
There was some suggestion that these provinces could provide a model for the poorerperforming provinces and interaction during the project would increase average
performance of participating provinces. Apart from some study tours to insuranceimplementing districts in Bali, there appeared to be limited interaction between
provinces.
142.
If the pairing of higher- and lower-performing provinces was done with the aim
of lifting average performance, a number of activities such as institutional twinning,
reciprocal work placement, and a broader program of information and experience
exchange should have been included in the design. In addition to a vague rationale for
province selection, the allocation of funds among participating districts was not based
on evidence of poverty prevalence or health performance indicators. Analysis provided
in the relevance section of this report suggests that district allocation was not propoor.
2.

District capacity and monitoring systems

143.
Insufficient district capacity and internal administration and management
systems constrained the success of health service delivery, and still do. Project delays
reflected cumbersome procedures in procurement, contract review, disbursement,
liquidation, and approval processes. The work of technical and administrative units was
not adequately synchronized for efficient project implementation. Benefit monitoring
and evaluation was not integrated. Many of the indicators established to track
institutional development were not specific, and required definition from a time-bound
and quality-standard point of view. In a decentralized service system, local health

The allocation
of funds was
was
not based on
evidence of
poverty
prevalence or
health
performance
indicators
Project delays
reflected
cumbersome
procedures

38
Measurable
and timetimedependent
indicators
should have
been used

Decentralized Health Services Project in Indonesia


departments need a functioning M&E system to track investment activities and outputs,
and health sector performance and impact. Technical monitoring guidelines such as the
World Health Organization's handbook45 have been developed to monitor and evaluate
health system performance. Measurable and time-dependent indicators, such as those
outlined in the handbook, should have been used to monitor and evaluate project
performance and consolidate in project completion and implementation reporting.
3.

Project design and scheduling

144.
Overall the project design was understood and large-value components were
successfully implemented, despite delays in progress. The use of a project loan coupled
with TA was the proper modality given the limited sector assessment and considerable
policy uncertainty at the time of decentralization. Reviews of other ADB loans, such as
the performance evaluation report for the Punjab Devolved Social Services Program,46
noted that project lending is best suited to demarcate specific capacity development
measures needed in the implementation of agreed reforms [as this] is a less predictable
process that requires some flexibility and responsiveness. The DHS design gave
flexibility to adapt to the evolving decentralization process. It contributed to the overall
capacity building of human resources with initiatives such as fellowship programs,
resulting in the placement of trained staff into higher management positions in project
areas. Districts planning capacity and vision of the proposed project activities was not
sufficient during project startup and early review missions noted this problem.

The design
was also
overambitious

145.
The implementation schedule did not accommodate an appropriate time frame
or activities to foster early project ownership. The design was also overambitious in its
geographic scope. More than 70 districts were involved in implementation, with
differing levels of capacity, high staff turnover, and all with limited experience in
international project operations. The number of provinces should have been reduced
and resource allocation should have been based on poverty and health status.

DistrictDistrictspecific plans,
plans,
supported by
more
substantive
TA, should
have been
developed

146.
District-specific plans, supported by more substantive TA, should have been
developed. Procurement could have been packaged into a smaller number of largevalue contracts, reducing the delays associated with ADB and government procurement
processes. Concentration of funding in a smaller geographic area would also increase
the probability of provincial health indicators reflecting positive outcomes of the
project. As already noted, project investment accounted for less than 2% of district
expenditure, which limits the scope for funding to have a significant impact on the
regional health statistics (para. 63). Larger allocations per district would not have had a
substantial impact given the small size of the project relative to overall district health
expenditure.

B.

Lessons

147.

Proper problem analysis is needed based on risk and mitigation analysis before
rolling out a program under decentralization. Problem analysis47 was undertaken to
some degree during project preparations. Risks were outlined in the project framework
(under Assumptions and Risks) but little detail was given in the main text of the report

45

46

47

World Health Organization. 2010. Monitoring the building blocks of health systems: a
handbook of indicators and their measurement strategies. Geneva.
ADB. 2012. Performance Evaluation Report: Punjab Devolved Social Services Program in
Pakistan. Manila.
ADB. 2007. Guidelines for Preparing a Design and Monitoring Framework. Manila.

Issues, Lessons, and Follow-Up Actions


and recommendation of the President, or in the project design. Greater attention
should have been placed on the political economy of decentralization, such as outlining
key stakeholders, defining differential impacts of reforms, and identifying risks
associated with possible future behavior of key stakeholders,48 along with capacity
assessments across the broad geographic spread of the project. It would have been
important to better assess the factors that help or hinder managerial and operational
institutionalization of local health systems. The project did not do so. As a result, the
standardized approach may not have worked in all districts. Decentralization is an
evolving process and requires time to develop adequate capacity.49

148. Results from successful pilot initiatives should be disseminated in time and
used in scaling up such initiatives to maximize the impact. The health insurance scheme
piloted in the Jembrana district of Bali has been recognized as one of the successful
initiatives and is now being rolled out across the entire province. A number of lessons
have emerged for ADB and the World Bank from piloting of this reform. This view was
shared by people interviewed during the evaluation and by the World Bank, who stated
that more lessons could undoubtedly be drawn by undertaking a comparative study of
all decentralization experiences, including those supported by development partners
like the World Bank and ADB. Such a study could focus on the different ways provinces
have established the province-district relationship, the relative costs and benefits of
each, and the variety of ways districts and provinces have developed to exercise
autonomy in the health sector despite continuing lack of clarity in the policy
environment and their dependency on central government funding (footnote 12).
149.
The scope and implementation
implementation coverage of projects should be manageable.
For significant impact, project investment should not be spread too thinly nor be so
ambitious in scope as to hamper implementation. The project interventions, while
generating some positive outcomes locally, did not show much impact at the provincial
level because they were too small to produce a broader impact. Concentrating
investment in a few provinces would have been more effective. Strategic investment
choices based on poverty and health needs should have been made when deciding
geographical coverage.
150.
Capacity development requires continued commitment. Indonesia still has a
relatively high MMR (para. 111). Capacity development in the areas of obstetric and
neonatal care is still required, and the evaluation team came to know from the district
health departments that they have limited finance to meet this capacity development
requirement. According to key informants, basic emergency obstetric and neonatal care
centers are often not used to assist with complications due to perceived lack of
capability, and that not all districts have a functioning hospital with required
emergency capability. Considerable further investment in infrastructure and human
resource capacity will be required to meet the maternal mortality MDG target by 2015.

48
49

ADB. 2013. Guidance Note: Use of Political Economy Analysis for ADB Operations. Manila.
In reviewing the World Bank provincial health projects, it was noted that no one could see
clearly how the implementation of decentralization would evolve; the important thing was to
put mechanisms in place which could respond well to opportunities and risks further
analysis would have been helpful looking at how decentralization could be expected to impact
the underlying causes of the entrenched problems in the sector, as some problems could be
transformed by decentralization and those which would likely be unaffected by
decentralization (footnote 12). The evaluation team concurs with this finding. Strengthweakness-opportunity-threat and risk analysis should have been coupled with strategies to
mitigate these risks.

39

40

Decentralized Health Services Project in Indonesia


151.
health--care services.
Greater synergy is required between family planning and health
Inclusion of family planning advocacy in future programs could help reduce maternal
mortality, as high fertility and maternal mortality are correlated. During key informant
interviews, many expressed an opinion that more support for family planning is
needed. It will require a coordinated and balanced joint approach from MOH and
BKKBN, supported by appropriate resources.
152.
A comprehensive approach is needed for a wellwell-functioning health system.
Well-functioning provincial and district systems have proven to contribute to overall
coordination and control of local health service delivery. During evaluation, health
facility staff frequently cited the development of human resource capacity through
longer-term fellowships as a key benefit of the project. Many of the project-supported
fellows are now in key positions within the district and provincial health services. Many
fellows thought that their careers would not have progressed so well without the
project. The skills for networking, problem-solving, and data analysis and presentation
that these fellowships help develop are still being used.
153.
ADB has helped improve infrastructure and promote reforms in a
comprehensive manner. This was seen as a key positive feature of the project.
Informants thought that by combining better planning and infrastructure and higherquality human resources, access to health services and their quality had improved.
Many felt that without a comprehensive approach this might not have happened.
154.
A good M&E system can help develop a better understanding of what works
and what does not.
not While the potential benefits of a comprehensive approach were
noted during the evaluation teams field visits, there is limited effort to quantify these
benefits systematically. A survey of stakeholders was conducted at project exit and
documented in a benefit-monitoring report. A range of customer satisfaction questions
were posed and scored in the survey. Satisfaction indicators related to benefits
associated with (i) training for midwives, (ii) utilization of skills by staff after training,
(iii) availability and equipment to support MNCH, (iv) quality of civil works,
(v) operations research, (vi) problem-solving ability of district health officers, (vii) staff
motivation, (viii) political commitment, (ix) effectiveness of health reforms, (x) level of
district funding (xi) health center utilization, (xii) safe delivery, and (xiii) patient
satisfaction. However, no baseline survey was conducted with the satisfaction survey,
so changes in with-and-without or before-and-after comparisons cannot be assessed.
Similarly, no non-project facility appears to have been included in the survey.
This would have helped in determining incremental benefits from project intervention
compared with non-project situations.
155.
Generally accepted health systems strengthening indicators, such as those
outlined in the World Health Organization's handbook (footnote 45), should be
included in a national and project-related M&E system. To determine project impact,
a focused survey of health status and client attitudes in project and non-project
provinces should have been conducted at the start (to establish a baseline) and at
project completion. Without this information it is not possible to attribute and quantify
shorter-term project benefits.
156.
Proper evaluation
evaluation of capacity development would help assess the utility of
current investment and justify further investment in human resource development.
A large number of staff in health service delivery received training or fellowships under
the project. However, there has not been any attempt to document actual benefits
stemming from this investment. A proper analysis will help the central, provincial,

Issues, Lessons, and Follow-Up Actions


and district governments to determine skills gaps that can be targeted over the next
couple of decades. Periodic evaluation of the value and impact of expenditure on
capacity development should have been conducted. Although the number of physicians
and the ratio of physicians to population have increased in all provinces and in rural
areas, deployment practices and inequitable distribution remain serious concerns. M&E
of current deployment and a skills audit are crucial to ascertain whether an effective
long-term health workforce strategy is being implemented (footnote 9).

C.

FollowFollow-up Actions

157.
The ADB-supported DHS2, approved in 2003 (footnote 13) and ongoing,
has incorporated some of the lessons derived from the implementation of this project.
The provinces covered in the DHS2 are different from those in the DHS. Nevertheless,
this evaluation has that are applicable to the ongoing project as well as any future
investment by ADB, other development partners and the government. The evaluation
has four key recommendations:
158.
ADB should remain engaged with the Government of Indonesia in the health
sector since there is tremendous need and scope for supporting health services and
health policy reforms to achieve better health outcomes for more people, including
some of the key MDG indicators such as maternal mortality and infectious disease
control. The country is still lagging in these areas and much of the current donor
support is associated with the control of particular infectious diseases rather than
developing the health system more broadly. Although health systems strengthening
has gained more focus in current donor support, a multilateral financing institution
such as ADB has a comparative advantage in encouraging the government to mobilize
resources for the health sector, assisting public financial management and tax reform,
helping formulate health insurance, and creating an environment conducive for private
investment.50
159.
The government should provide a clear funding framework and simplified
financing arrangements for enhancing local, district, provincial, and central health and
family planning services. Local health financing in Indonesia is characterized by
complicated budgeting and planning cycles, which involve a multitude of financing
channels. Consequently transaction costs are high, and delays in fund disbursement
effectively reduce the window in which health facilities can operate efficiently within a
given year. It is vital to simplify these procedures and funding arrangements, and give
greater autonomy to districts over operational budgeting and human resources. One of
several options may be to provide matching grants (1:1) to local, district, and provincial
entities. More stable funding arrangements will most likely help them plan and
implement health activities.
160.
Provincial governments, with support from central government, should
promote twinning arrangements with other health facilities within and outside their
provinces for human resource development in the health sector. This would require a
credible assessment of current deployment, and a skills audit of health service delivery
facilities. Where possible, efforts should be made to encourage intra-district or interdistrict mobility of experienced staff to go on secondment arrangements, supported by
a genuine incentive scheme.

50

ADB. 2013. Health in the Post-2015 Development Agenda for Asia and the Pacific. Manila.

41

42

Decentralized Health Services Project in Indonesia


161.
BAPPENAS, MOH, BKKBN and the National Statistics Office need to take a joint
lead role in ensuring that healthhealth-related statistics are collected regularly and records are
updated on time
time so that information and data can be extracted for policy formulation
and program development. This would be possible if each agency ensured that flow of
data and information from local, district and provincial health facilities is uninterrupted
and consolidated efficiently. The government may seek assistance from one of the
development partners (including ADB) if funding emerges as a key constraint.
BAPPENAS should also organize an annual event where experiences and lessons from
various pilot initiatives are shared across national and international agencies involved in
promoting better health outcomes.

Appendixes

APPENDIX 1: DESIGN AND MONITORING FRAMEWORK

Design Summary
Goal
Improved health
status of the
population in all
project districts

Purpose
Improved health
and family
planning services
in the project area

Performance Targets/Indicators
Actual
Appraisal
(Average Provincial Figures)
Appraisal Target
By 2010, in the
project districts
MMR: 200
deaths per
100,000 live
births, or at least
30% lower than
local benchmark
data

IMR: 30 deaths
per 1,000 live
births, or at least
30% lower than
local benchmark
data

IMR in the original project


provinces fell from 42 deaths
per 1,000 live births in 2003 to
36.6 in 2010. This represents a
reduction of 13%. Increases in
IMR were reported in Bali, North
Sulawesi, and Central Sulawesi
between 2003 and 2010.

U5MR: 40 death
per 1,000 live
births, or at least
30% lower than
local benchmark
data

U5MR fell from 59.7 per 1,000


live births in 2003 to 43.3 in
2010. This represents a decrease
of 27%. The largest decrease
was reported for Bali, from 30.3
to 5.4 deaths per 1,000 births

LEB 70 years, or
at least 2 years
more than local
benchmark data

The PCR reported that Riau, Bali,


and North Sulawesi had attained
a life expectancy of at least 70
years

District targets for at


least 47 of the
62 project districts or
cities (75%) met.

Access of the poor


to essential health
and family
planning services
guaranteed

According to government
statistics, MMR decreased from
232 deaths per 100,000 live
births in 2003 to 84 in 2012.
This is equivalent to a 63%
decrease. The largest decreases
were evident in Southeast
Sulawesi and NAD, where
reductions of more than 200 per
100,000 live births were
recorded.

Contact rates at
least 25% higher
than benchmark
data

Data Sources/Reporting
Mechanisms

The contact rate increased by


31% between 2003 and 2010
for outpatient services. Increases
were not evident in all
provinces. Decreases were
apparent in Southeast Sulawesi
and Bali, while large increases
(between 61% and 360%) were
recorded in Central Sulawesi,
North Sulawesi and Riau.

Annual district health


profiles
UNICEF and WHO data
or estimates

National
Socioeconomic survey
health module
Demographic and
Health Survey
BKKBN reports
Local government
budget

Design and Monitoring Framework

Design Summary

Local capacity to
develop and
maintain quality
health systems
and healthhealth-care
delivery services

Performance Targets/Indicators
Actual
Appraisal
(Average Provincial Figures)
Appraisal Target

Unmet needs for


Unmet demand for family
family planning
planning remained relatively
30% lower than
constant between 2003 and
benchmark data
2007, and then increased in
2010. Unmet demand has
increased most notably (by more
than 5%) in North Sulawesi,
Central Sulawesi, and Southeast
Sulawesi between 2003 and
2010.

Data Sources/Reporting
Mechanisms

Sustainability:
ratio of local
governments
budget for
health over total
public
expenditure at
least 4%

The average ratio in 2002 was


3.6% and increased to 9.8% by
2011. The largest increases were
in the provinces of Central
Sulawesi, NAD, and Southeast
Sulawesi. The latter's ratio
increased from 1.7% to 19.7%.
Only aggregate provincial ratios
were provided so it is unclear
how many districts have made
the threshold.

Sustainable local
health safety-net
program
No breakdown
of health care
services delivery

Indicator for this target unclear

Indicator for this target unclear

Specific
indicators:
availability of
essential drugs
and
contraceptives

Average contraceptive
prevalence rate among married
women (15-49 years) remained
stable across four project
provinces between 2003 and
2010. Increases were apparent
in Riau, Bali and Southeast
Sulawesi.

Targeting public
subsidies to the
poor: local
health safety net
program; local
primary health
care services

The PCR found that visit rates


among the poorest 20%
increased by at least 25%
between 2001 and 2007. World
Bank project reviews also found
the poor using health facilities
to a greater extent in target
provinces.

Awareness of
gender-specific
health needs

Quality of district
proposals

The PCR noted that the number


of midwives trained in lifesaving
skills increased by at least 50%,
except in NAD and North
Sulawesi. Numbers of midwives
and nurses in each province
dramatically increased between
2003 and 2010. For example,

45

Performance
monitoring system;
field visits; NGOs and
media reports
Experts assessment of
local health safety-net
program; NGOs and
media reports
Local reproductive
health services
Experts assessment of
the proposals and
reports

46

Decentralized Health Services Project in Indonesia

Design Summary

Health sector
reforms

Health services
infrastructure

Project
management

Performance Targets/Indicators
Actual
Appraisal
(Average Provincial Figures)
Appraisal Target
the number of midwives in NAD
Quality of district
increased from 325 to 8,930.
yearly
implementation
reports

This is a non-specific indicator


and should have been specified
on a time-bound and minimumquality basis. The PCR noted
that proposals and annual work
plans were reviewed, but there
was no clear guidance on
reporting implementation
progress.
Local health

These outcomes were not


sector reforms
routinely measured as part of
introduced:
project M&E. Discussions during
increased
the evaluation indicated that
efficiency,
local health reforms
increased quality
strengthened health care
in health sector
management capacity and
service delivery. Case studies are
provided in the main text of the
report.
Physical health

The number of health centers in


infrastructure:
project provinces more than
clean and
doubled between 2003 and
rehabilitated
2010. Large increases were
health facilities;
evident in NAD and Southeast
equipped with
Sulawesi. Much of the increase
required medical
in NAD was not related to the
equipment
project. The project was
responsible for 188 health
centers with beds and 162
health centers without beds,
and renovation of 158 subdistrict health centers. Hospital
beds increased by 54% between
2003 and 2010 in project
provinces. A total of
86 hospitals were renovated and
seven were constructed using
project funds. Medical
equipment and ambulances
were procured to support
construction and health needs in
each district.
Timely

Project implementation was


implementation
delayed by late disbursement of
of central-level
counterpart funds and by
project activities
planning and implementation
(100% MOHSW
bottlenecks.
and BKKBN
proposals) and
completion of
district proposals
in at least 50 of
the 62 project

Data Sources/Reporting
Mechanisms

Experts assessment of
the need and quality of
local reforms

Project reports
Field visits

Project reports
Field visits
Project completion
report

Design and Monitoring Framework

47

Performance Targets/Indicators
Actual
Data Sources/Reporting
Design Summary
Appraisal
(Average Provincial Figures)
Mechanisms
Appraisal Target
districts
Training activities
ADB: $16,898.30
ADB: $19,977.0

Withdrawal
Counterpart funds: Counterpart funds: $12,685.0
applications
$4,592.80

Disbursements
Consulting
ADB: $3,107.30
ADB: $833,205.0

Annual audit reports


services
Counterpart funds: Counterpart funds: $1,139.0
and financial
$3,186.20
statements
Civil works
ADB: $12,236.30
ADB: $12,707.7
Counterpart funds: Counterpart funds: $6,716
$5,677.10
Equipment and
ADB: $17,243.10
ADB: $17,259.0
materials
Counterpart funds: Counterpart funds: $3,928
$1,067.00
Operation studies
ADB: $1,170.50
ADB: $1,363.8
Counterpart funds: Counterpart funds: $963
$713.70
Project
ADB: $3,537.80
ADB: $5,192.5
management
Counterpart funds: Counterpart funds: $6,519.9
$4,383.30
BKKBN = National Population and Family Planning Board, IMR = infant mortality rate, LEB = life expectancy at birth, M&E
= monitoring and evaluation, MMR = maternal mortality ratio, MOHSW = Ministry of Health and Social Welfare, NAD =
Nanggroe Aceh Darussalam, NGO = nongovernment organization, PCR = project completion report, U5MR = under-5 mortality
rate, UNICEF = United Nations Childrens Fund, WHO = World Health Organization.
Source: Independent Evaluation Department and project documents.

APPENDIX 2: RATING MATRIX FOR CORE EVALUATION CRITERIA


Table A2.1: Rating Matrix for Project Performance Evaluation Criteria
Parameter/
Criteria
Relevance
Effectiveness in
Achieving
Outcomes
Efficiency in
Achieving Outputs
and Outcomes
Preliminary
Assessment of
Sustainability
Overall
Assessment
Impact

ADB Performance
Borrower
Borrower
Performance
Technical
Assistance

PCR
Relevant
Effective

IED
Relevant
Effective

Less Efficient

Less than
Efficient

Likely
Sustainable

Likely
Sustainable

Successful

Successful

Not Rated

Limited

Satisfactory
Satisfactory

Quality of PCR

Satisfactory
Less than
Satisfactory
Less than
Satisfactory

Reason for Disagreement with PCR Rating /


Comment

No significant differences exist between project


and not-project districts or provinces. While
access has improved, actual impact remains
limited.
Delay in budget release caused implementation
delays.
Evidence of consultants outputs could not be
substantiated. Only part of overall outputs could
be used by the project. Coordination of individual
consultants output remained a challenge.

Satisfactory

IED = Independent Evaluation Department, PCR = project completion report.


Source: IED. 2006. Guidelines for Preparing Performance Evaluation Reports for Public Sector Operations. Manila.

Appendix A2.2: Core Evaluation Criteria Rating Values


Rating Value
3
2
1
0

Relevance
Highly Relevant
Relevant
Less than
Relevant
Irrelevant

Effectiveness
Highly Effective
Effective
Less than
Effective
Ineffective

Efficiency
Highly Efficient
Efficient
Less than
Efficient
Inefficient

Sustainability
Most Likely
Likely
Less Likely
Not Likely

Source: IED. 2006. Guidelines for Preparing Performance Evaluation Reports for Public Sector Operations. Manila.

APPENDIX 3: ECONOMIC REEVALUATION


A.

Introduction

1.
The project is reevaluated following the Asian Development Banks (ADB) Guidelines for the
Economic Analysis of Health Sector Projects.1 The economic life of the project was assumed to be
20 years from the commissioning of health infrastructure, a discount rate of 12% was applied, and cost
and benefits were included in constant 2013 prices. The goal of the project was to help improve the
health status of populations in target provinces, by developing human resource capacity, introducing
health reforms to improve the coverage of services, and developing health infrastructure to improve
access. These activities generate health benefits by reducing the direct costs associated with treating
disease, along with the indirect costs of impaired productivity and premature mortality. These
economic benefits were quantified through a comparison of with and without project situations.
2.
Following cost-benefit analysis conducted at appraisal, economic benefit estimation was
conducted using a disability-adjusted life year (DALY) approach to measure the potential changes in
burden of disease after project implementation. DALYs measure the number of years of life lost (YLL)
due to premature mortality and quality of life through disability weights for differing diseases.
The project was expected to reduce the gap in burden of disease, as measured by DALYs, between
project provinces and developed countries.
3.
At the time of implementation Indonesia had a far higher burden of disease per person than
developed (Organisation for Economic Co-operation and Development [OECD]) countries, which is
referred to as the DALY gap. This gap was estimated to have narrowed, using assumptions associated
with the 2010 Global Burden of Disease Study.2
4.
The number of DALYs gained (through a reduction in the burden of disease) was multiplied by
the average earnings in project provinces. At appraisal, these benefits over 10 years were compared
with the costs of project activities, with resulting benefit-cost ratios varying from 11.8 in Riau to 1.1 in
Bengkulu. The overall benefit-cost ratio for the project was 4.5.
5.
A result of this order indicates that for each dollar of project investment, the project generates
$4.5 of benefits. The rationale for such a high return was not clearly stated in appraisal documents, so
the analysis is redone in this section using the comparison of service delivery and health outcomes for
project and non-project provinces.

B.

Indonesian Burden of Disease

6.
Infant and maternal mortality rates have fallen in Indonesia. Maternal mortality ratios show a
declining trend but are high relative to other East Asia countries. It is evident in Table A3.1 that
Indonesia has high maternal mortality when compared with Thailand and the Philippines, but lower
ratios than Cambodia and the Lao Peoples Democratic Republic. Maternal mortality is caused by
problems such as postpartum and antepartum hemorrhagewhere too much blood is lostalong
with placental abnormalities, infection, and unsafe abortion.

1
2

ADB. 2000. Handbook for the Economic Analysis of Health Sector Projects. Manila.
Institute for Health Metrics and Evaluation, Human Development Network. 2013. The Global Burden of Disease:
Generating Evidence, Guiding Policy: East Asia and Pacific Regional Edition. Seattle: World Bank.

50

Decentralized Health Services Project in Indonesia


Table A3.1: Maternal Mortality
Mortality
(per 100,000 live births)

OECD
Viet Nam
Indonesia
Philippines
Cambodia
Lao PDR
Thailand

1995
26
160
420
140
750
1200
54

2000
25
100
340
120
510
870
66

2005
19
74
270
110
340
650
54

2010
19
59
220
99
250
470
48

Change
(%)
(28)
(63)
(48)
(29)
(67)
(61)
(11)

Lao PDR = Lao Peoples Democratic Republic, OECD = Organisation for Economic Co-operation and Development.
Source: World Bank development Indicators.

Table A3.2: Infant Mortality


(per 1,000 live births)

OECD
Viet Nam
Indonesia
Philippines
Cambodia
Lao PDR
Thailand

1995
13
30
51
34
88
98
24

2000
10
25
41
30
82
85
19

2005
8
21
34
28
52
71
15

2010
7
19
28
25
37
58
12

Change
(%)
(48)
(37)
(45)
(26)
(58)
(41)
(50)

Lao PDR = Lao Peoples Democratic Republic, OECD = Organisation for Economic Co-operation and Development.
Source: World Bank development Indicators.

7.
Infant mortality has been decreasing. Mortality rates do not, however, provide a complete
indication of the health status of a population. The burden-of-disease metric referred to as the DALYs is
most often used to capture mortality and morbidity impacts of different diseases. One DALY represents
the loss of the equivalent of 1 year of full health. The loss of 1 year of life through premature mortality
would be estimated to be the loss of one DALY, denoted as YLL. Years of life can also be led with
disability inflicted by disease, known as years lived with disability (YLD). Weights have been established
for differing causes of disease to account for this morbidity. The YLL and YLD attributable to differing
diseases were first estimated as part of the 1993 World Development Report.3 This burden of disease
was updated in 2010. Burden of disease estimates for Indonesia and high-income countries are in
Table A3.3.
Table A3.3: DisabilityDisability-Adjusted Life Years per 1,000
1995
Indonesia
High-income countries
Gap
Annual gap reduction

375.0
275.2
99.8
-

Indonesia
High-income countries
Gap
Annual gap reduction

104.2
119.1
(14.9)
-

2000
DALY rate per 1,000
350.7
266.6
84.1
15.7
YLD rate per 1,000
104.6
120.9
(16.3)
1.4

World Bank. 1993. World Development Report. Washington, DC.

2005

2010

338.0
260.8
77.2
6.9

320.5
261.3
59.3
17.9

103.8
123.1
(19.3)
2.9

105.4
125.7
(20.4)
1.1

Economic Reevaluation

1995
Indonesia
High-income countries
Gap
Annual Gap Reduction

270.8
156.1
114.7
-

2000
YLL rate per 1,000
246.7
145.7
101.0
13.7

51

2005

2010

234.2
137.7
96.5
4.6

215.2
135.5
79.6
16.8

( ) = negative, = not available, DALY = disability-adjusted life year, YLD = years lived with disability, YLL = years of life lost.
Source: Institute for Health Metrics and Evaluation, Human Development Network. 2013. The Global Burden of Disease:
Generating Evidence, Guiding Policy: East Asia and Pacific Regional Edition. Seattle: World Bank.

8.
Assumptions underpinning the calculation of the reduction in the burden of disease gap were
not included in the appraisal document for the Decentralized Health Services Project (DHS), but the
approach was referred to in the follow-on Second Decentralized Health Services Project (DHS2). In the
DHS2, the economic analysis conducted at appraisal indicated that the gap in DALYs between nonIndian and Chinese Asian countries and established market economies was about 143 per 1,000
population. The gap in DALYs was estimated by disaggregating the major disease categories, using
relatively recent burden of disease data from previous health projects in the 1990s. Estimates in the
global burden of disease study of 2010 update this estimate to a gap of 84.1 DALYs per 1,000 in 2000.
This base difference is used in the reevaluation of the projects economic analysis. It should also be
noted that the decrease in burden-of-disease gap is estimated to be associated with reductions in YLL,
rather than YLD.

C.

Project Contribution to Health Financing

9.
Indonesias health spending levels and trends are contained in Indonesias National Health
Accounts. Health financing is very fragmented with numerous channels of public financing from the
central to lower governments, along with insurance, private spending, and local government support.
Before 2004, private health expenditure had the largest share of total health spending and averaged
58%, but by 2006 private and public expenditure were about equal.
10.
Total health expenditure at the district level was estimated to be around $38 per person.
Multiplied by the total population in the project implementation area, total district health expenditure
was around $700 million at the time of project implementation. The total cost of the project estimated
in the report and recommendation of the President was $87.03 million, with a realized expenditure by
June 2008 of $84.60 million.
11.
Averaged over a 7-year implementation period, project expenditure was around 1%-2% of total
district spending. When making assumptions about the potential impact of a project loan on provincial
outcomes, the relative contribution of the development partner as opposed to other sources of finance
needs to be considered. In this case, it is evident that development partner financing comprises a very
small part of overall financing at the district level.

D.

Major Assumptions

12.
The DHS had the objectives of improving the treatment of diseases and quality of family
planning services, and injury-related clinical services. This was estimated in the report and
recommendation of the President's economic evaluation to reduce the DALY difference between OECD
market economies and Indonesia by 4%. Based on a gap of 84.1 DALYs per 1,000 inhabitants, a gap
reduction of this order would have lowered the burden of disease in Indonesian target provinces by
3.4 DALYs per 1,000 inhabitants.

52

Decentralized Health Services Project in Indonesia


13.
Studies in the Asia and Pacific region of factors constraining the use of maternal and child
health services found that the financial costs of treatment were the dominant barrier to obtaining
medical care in Bangladesh, Cambodia, and Pakistan. Distance and transport difficulties were reported
as major factors, so better access as a result of infrastructure developed under the project and
improved contact rates are likely to have had some impact on the health of those in project provinces.4
14.
Because there was no substantial difference in the health outcomes of the project and nonproject provinces (with and without scenarios) the reduced burden-of-disease gap attributable to
the project is smaller than the estimated impact at appraisal of 4%. Informants during the evaluation
were adamant, however, that the project had positive health benefits. While the loan resources
comprise less than 2% of the district health expenditures, it was noted that those resources are used
for critical activities such as staff training and investments in remote areas, which are critical for health
outcomes. Health centers constructed by the project are intended to treat most health conditions, and
the training of staff has improved the quality of health services beyond the mother, newborn, and child
health (MNCH) services. Management staff also has been trained to make better use of the scarce
resources. Although a significantly reduced burden of disease cannot be quantified using Indonesia
Demographic and Health Survey data, a breakeven burden of disease required for the project to break
even is determined in the next section.

E.

CostCost-Benefit Analysis

15.
The quantitative cost-benefit analysis for the project follows the methodologies outlined in
ADBs Guidelines for the Economic Analysis of Health Projects and a World Bank publication.5
The analysis covers a projection period of 20 years and is in constant 2013 dollar prices. Incremental
recurrent costs are estimated at 5% of investment costs and projected over the cost-benefit analysis.
An annual minimum wage of $1,389 is used for base calculations.
16.
The main benefits that were quantified include health benefits from reducing the burden of
disease in Indonesia as a result of capacity and infrastructure development. A DALY gap of 84 per 1,000
inhabitants is included to determine the breakeven gap reduction required to generate a positive
economic benefit. This gain in life years could also be valued at the annual minimum wage in
Indonesia. The indirect cost estimation draws on the human capital approach, which has been widely
used to assess the productivity losses from illness or injury as measured by income forgone due to
morbidity, disability, and mortality. Labor force participation rates and earnings of affected individuals
are used to calculate the value of productivity losses due to morbidity and premature mortality.
An annual wage of $1,389 is used and adjusted using a shadow wage factor of 0.5, because not all
DALYs are associated with working-age people.
17.
The World Health Organizations Commission on Macroeconomics and Health evaluated the
links between health and poverty.6 Analysis undertaken as part of the commission estimated that by
averting 8 million deaths, around $180 billion in economic benefits would be generated per year, or
$22,500 per death in direct economic savings by 2015. The estimate assumed that each DALY is valued
at 1 year of average per capita income. Gross national index per capita was $3,420 in 2012.7 This cost is
4

R.P. Rannan-Eliya et al. 2012. Impact of out-of-pocket expenditures on families and barriers to use of maternal

and child health services in Asia and the Pacific: Evidence from national household surveys of health care use and
expenditures. Summary technical report. Manila: ADB.
5

E. Bloom and P. Choynowski. 2003. Economic Analysis of Health Projects: A Case Study in Cambodia, ERD
Technical Note No. 6, Manila: ADB; and P. A. Musgrove. 2003. Health Economics in Development. World Bank,
Washington, DC, November 2003 issue.
WHO. 2001. Macroeconomics and Health: Investing in Health for Economic Development. Report of the
Commission on Macroeconomics and Health. Geneva.
Gross national index per capita (formerly GNP per capita) is the gross national income, converted to U.S. dollars
using the World Bank Atlas method.

Economic Reevaluation

53

used as a high bound estimate in sensitivity analysis. A lower bound of around $1,075 per year
(the minimum was in Central Sulawesi) is included to gain an appreciation for the robustness of results
to changes in value of labor assumptions.

F.

Evaluation Results

18.
The breakeven burden-of-disease reduction is estimated to be 2%, and leads to the reduction
of nearly 0.04 million DALYs per year. A marginal decrease in burden of disease of this order is possible
despite provincial health indicators not demonstrating clear impacts. As noted in the main text
(para. 108), many Indonesia Demographic and Health Survey indicators are a moving average and
recent estimates may not reflect the health benefits of the project.
Table A3.4: CostCost-Benefit Analysis

Year
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
Total
EIRR
BCR
NPV

Economic
Cost
($ million)
2.99
11.19
21.35
24.33
11.46
9.90
12.51
10.85

104.59
104.59

12%

O&M
($ million)
0.10
0.31
0.77
1.45
1.79
2.02
2.27
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
58.01

Total
Cost
3.09
11.50
22.12
25.78
13.25
11.91
14.78
13.20
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
162.60

Population
in Target
Provinces
(million)
19.13
19.33
19.52
19.71
19.91
20.11
20.31
20.52
20.72
20.93
21.14
21.35
21.56
21.78
22.00
22.22
22.44
22.66
22.89
23.12
23.35
23.58
23.82
24.06
24.30
24.54
24.78
23.12

Reduction in
Disease
Burden
(DALYs
millions)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
0.03
0.03
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.76

Economic
Benefits
($ million)
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
24
24
24
25
25
25
25
26
26
26
26
27
27
27
28
28
28
28
29
27
525

Net
Benefits
($ million)
(3.09)
(11.50)
(22.12)
(25.78)
(13.25)
(11.91)
(14.78)
(13.20)
21.60
21.84
22.08
22.33
22.57
22.82
23.07
23.33
23.59
23.84
24.11
24.37
24.64
24.91
25.18
25.46
25.73
26.01
26.30
24.37
362.51
12.0%
1.00
0.00

( ) = negative, BCR = benefit-cost ratio, DALY = disability-adjusted life year, EIRR = economic internal rate of return, n.a. = not
available, NPV = net present value.
Source: Independent Evaluation Department computations.

54

Decentralized Health Services Project in Indonesia

G.

Sensitivity Analysis

19.
There is a substantial amount of uncertainty surrounding a number of variables used in the
baseline evaluation. The impact on investment returns resulting from changes in the value of labor is
provided in Table A3.5, and from changes to reductions in the burden-of-disease gap in Table A3.6.
Table A3.5: Sensitivity of Investment
Criteria to Changes in Assumed Wage Rate

Investment Criteria
Criteria
Net present value ($ million)
Benefit-cost ratio
Economic internal rate of return (%)

Low
$1,075

Base
$1,389

High
$3,420

(75.67)
0.77
9.2%

0.00
1.00
12.0%

489.47
2.46
23.2%

20.
It is calculated that the economic internal rate of return of the project would decrease by 2.8%
if the value of labor was reduced to $1,075 per year.
Table A3.6: Sensitivity of Investment
Criteria to Changes in Assumed Discount Rate

Investment Criteria
Net present value ($ million)
Benefit-cost ratio

Low
5%

Base
12%

200.06
1.95

0.00
1.00

High
20%

(284.61)
0.52

21.
It is calculated that the net present value would increase by $200 million if the discount rate is
decreased to 5.0%.

H.

Conclusions

22.
The economic benefits of undertaking the project are positive if a 20-year projection is
included. At appraisal it was estimated that the gap between Indonesia and OECD countries could be
reduced by 4, which would be a positive economic benefit if the burden of disease were reduced by
2%. Although no substantial changes were evident in health outcome indicators between project and
non-project provinces, a marginal increase is likely based on discussions in this report. Correspondingly,
the project is more than likely generated positive economic benefits.

APPENDIX 4: HEALTH OUTCOMES IN PROJECT PROVINCES


Table A4.1:
A4.1: Infant Mortality Rate in Project Provinces of Indonesia
(Deaths per 1,000 Live Births)
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Provincesa
Difference

1997
58.8
45.5
60.4
72.3
39.5
47.6
94.5
78.1
49.0
9.8

2003
42.1

2007
2007
38.5
25.0
37.0
46.0
34.0
35.0
60.0
41.0
65.6
(27.1)

43.0
53.0
14.0
25.0
52.0
67.0
43.2
(1.1)

2012
36.6
47.0
24.0
29.0
29.0
33.0
58.0
45.0
37.4
(0.7)

NAD = Nanggroe Aceh Darussalam.


Note: a Non-project provinces included all provinces except those covered by World Bank provincial health projects since 2000
(Yogyakarta, Lampung, North Sumatra, West Java, Banten, Jambi, East Kalimantan, West Kalimantan, West Sumatra)
and Jakarta. Population-weighted averages provided for project and non-project provinces.
Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2012). Non-project provinces do not include Jakarta,
or provinces covered by the World Bank's provincial health projects.

Table A4.2:
A4.2: Deliveries and Provision of Cesarean Section, By Province
(2003, 2006, 2010)
Deliveries (000)
Province
NAD
Riau
Bengkulu
Riau Islands
North Sulawesi
Central Sulawesi
Southeast Sulawesi
Bali

2003
54
79
63
n.a.
52
79
77
96

2006
76
64
77
81
63
72
72
84

2010
76
62
76
79
77
78
59
66

2003
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.

Cesarean Section
(% of Deliveries)
2006
8.0
8.4
3.7
1.9
2.3
3.2
10.7
1.7

2010
9.8
19.2
5.9
7.4
4.2
7.6
7.6
14.1

NAD = Nanggroe Aceh Darussalam, n.a. = not available.


Source: Provincial Health Offices, Indonesia.

Table A4.3:
A4.3: Percentage of Mothers Receiving Assistance during Delivery from a Skilled Provider
(%)
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Difference

1997
43.6
39.4
44.2
42.3
77.6
45.7
22.6
25.3
40.4
3.2

2003
69.1
74.0
68.6
87.8
85.7
54.0
42.0
68.1
1.0

2007
75.8
72.5
84.9
72.3
92.6
87.3
59.6
56.6
71.9
3.9

2012
83.9
89.8
86.4
87.2
98.7
85.8
62.9
65.9
82.1
1.9

NAD = Nanggroe Aceh Darussalam.


Note: Skilled providers include doctors, nurses, midwifes, and auxiliary nurses or midwifes.
Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2012). Non-project provinces do not include Jakarta,
or provinces covered by the World Bank's provincial health projects.

56

Decentralized Health Services Project in Indonesia


Table A4.4:
A4.4: Percentage of Women Receiving Antenatal Care from a Skilled Provider
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Difference

1997
88.5
85.2
86.0
86.7
97.5
95.6
82.5
89.2
91.6
(3.1 )

2003
90.4
90.3
91.9
97.7
96.8
82.4
84.9
92.0
(1.6 )

2007
93.1
89.2
93.2
93.7
98.8
95.9
90.7
91.3
93.1
0.0

2012
95.6
95.3
95.8
96.5
99.3
95.1
93.2
93.1
95.7
(0.1)

NAD = Nanggroe Aceh Darussalam.


Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2012). Non-project provinces do not include Jakarta,
or provinces covered by the World Bank's provincial health projects.

Table A4.5:
A4.5: Percentage
Percentage of Children Aged 12
1223 Months Who Received All Basic Vaccinations
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Difference

1997
58.3
45.5
60.4
72.3
39.5
47.6
94.5
78.1
60.1
(1.8)

2003
63.5
55.2
69.2
80.3
68.6
66.5
52.8
58.6
4.9

2007
52.4
26.8
41.4
54.9
72.2
76.1
50.3
64.6
61.9
(9.5)

2012
65.0
49.7
57.6
66.7
87.0
77.1
67.2
70.5
69.7
(4.7)

BCG = bacille Calmette-Gurin vaccine, DTP = diphtheria-tetanus-pertussis vaccine, NAD = Nanggroe Aceh Darussalam.
Note: Basic vaccinations include BCG, measles, and three doses each of DPT and polio vaccine (except polio 4).
Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2012). Non-project provinces do not include Jakarta,
or provinces covered by the World Bank's provincial health projects.

Table A4.6:
A4.6: Percentage of Children Taken for Treatment to a Health Facility by Mothers
Year
Project Provinces
Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
National

1997
%
53.3
69.3
48.2
45.6
63.4
51.7
48.7
51.7
54.8

NAD = Nanggroe Aceh Darussalam.


Source: Indonesia Demographic and Health Survey (1997, 2007, and 2012).

2007
%
57.1
68.9
49.7
47.8
89.3
58.4
42.9
41.4
51.0

2012
%
65.9
69.9
63.5
81.6
76.0
64.3
59.5
55.4
64.6

Health Outcomes in Project Provinces

57

Table A4.7:
A4.7: Percentage of Women Using Contraceptive Methods
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Difference

1997
55.0
37.1
48.0
66.6
68.1
71.2
51.7
53.1
58.2
(3.2)

2003
59.5
57.8
68.2
61.2
70.1
54.6
48.6
62.4
(3.0)

2007
61.1
47.4
56.7
74.0
69.4
69.3
63.6
50.7
61.9
(0.8)

2012
59.0
46.8
61.1
64.2
66.2
68.9
55.7
51.5
62.4
(3.3)

2007
8.5
12.0
9.1
6.1
5.8
6.1
8.3
12.9
15.4
(6.9)

2012
12.5
14.0
11.8
9.1
9.3
10.8
15.7
18.4
19.5
(7.0)

NAD = Nanggroe Aceh Darussalam.


Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2012).

Table A4.8:
A4.8: Unmet Need for Family Planning
Year
Project Provinces
NAD
Riau
Bengkulu
Bali
North Sulawesi
Central Sulawesi
Southeast Sulawesi
NonNon-Project Provinces
Difference

1997
8.6
10.3
12.7
7.4
5.8
4.4
9.4
8.9
15.5
(6.8)

2003
8.9
10.4
8.0
6.9
4.4
10.2
13.4
16.3
(7.3)

NAD = Nanggroe Aceh Darussalam.


Source: Indonesia Demographic and Health Survey (1997, 2003, 2007, and 2013).